Geological Survey Professional Paper 58 The Guadalupian Fauna

MOLLUSCOIDEA.

BRYOZOA.

The Guadalupian Bryozoa, so far as known, comprise 14 genera and 44 species, as indicated in the following list:

For the most part this group has proved rather surprisingly scanty, and my meager material would hardly have yielded so respectable an array of forms without the skillful manipulation of my friend Mr. Bassler, who made the sections which I studied and gave valuable aid in their investigation. The most abundant type is probably Acanthocladia guadalupensis, with the interesting group which I have placed under Domopora very nearly as plentiful. The greatest variety of species is found among the Fenestellas and Polyporas, which are more rare in the Guadalupe section itself than in the southern Delawares. Most of the types discriminated, however, are based on mere fragments. All the other forms, except possibly Fistulipora gradis var. guadalupensis, are rare.

In addition to being fragmentary, the Guadalupian Bryozoa proved unsatisfactory for study in another particular. A good deal of the material is silicified and thus unfitted for sectioning, while the structure was often found more or less obscured where siliceous replacement had not taken place.

In some cases the study of thin sections is less essential to the determination of species. Among the Domoporas I have discriminated species on external characters, and etched specimens, unless too coarsely silicified, were very favorable to this treatment. The Fenestellas, on the other hand, mostly proved to be too coarsely or too imperfectly replaced by silica to yield satisfactory results.

The Guadalupian Bryozoa differ considerably from those of the Salt Range. Many of the Indian forms which I should be disposed to place with the Bryozoa and compare with the present fauna Waagen and Wentzel have assigned to the cœlenterates. Such are the following.

Among the Bryozoa cited are the genera named below:

This list must be further modified by replacing Geinitzella by Batostomella, of which Waagen's genus is probably a synonym, and enlarging Fistulipora so as to include and eliminate Dybowskiella.

The two faunas have in common the genera Fistulipora, Stenopora, Fenestella, Polypora, Phyllopora?, Acanthocladia, Thamniscus, Rhombopora, and Goniocladia; while the Guadalupian contains species of Domopora, Meekopora, Leioclema, and Actinotrypa not found in the Salt Range, and the latter fauna contains Monotrypa, Orbipora, Batostomella, Hexagonella, and Synocladia not found in the Guadalupian.

There are thus a considerable number of generic types which are peculiar to each. Of these, foremost in importance on the part of the Guadalupian are without doubt the Domoporas, which form so abundant and striking a feature of that fauna. There appears to be no single genus which plays the same role in the Salt Range fauna, but Batostomella (Geinitzella) and Synocladia are somewhat important, and Hexagonella is sufficiently abundant and striking to deserve special mention, while Phyllopora and Thamniscus are rare in the Guadalupian and more or less unsatisfactorily identified.

Among the points of resemblance represented by the possession of types in common Goniocladia is perhaps the most noteworthy, for that genus, first described from Scotland, was for a long time known elsewhere only in India. Later it was obtained in the Guadalupe Mountains, and about the same time specimens were brought back from Alaska. Acanthocladia is to a less degree important, but aside from these the bryozoan genera which are common to the two faunas are such as have a wide dispersion and a long range and might be contained in almost any two faunas of late Carboniferous age.

Some of the species are also related, but the types are not so peculiar or the affinity so intimate as to be especially noteworthy. Perhaps the most important instance of specific relationship is to be found in Fistulipora grandis var. guadalupensis, which represents the type of fistuliporoid for which Waagen and Wentzel introduced the generic term Dybowskiella, and is closely allied to Dybowskiella or Fistulipora gradis. One may remark, on the other hand, in this connection the large development of Polypora in comparison with Fenestella, rather the reverse of what is indicated in the American fauna.

On the whole, therefore, the bryozoan faunas of the Salt Range and Guadalupian do not seem to me to indicate any marked relationship.

But scanty mention of Bryozoa has been found in literature dealing with the Himalayan region. Diener records a species of Fenestella in his first paper on Chitichun No. 1, and in that on anthracolithic fossils from Kashmir and Spiti two species of Fenestella, one of Protoretipora, and one of Acanthocladia. In a subsequent paper on the Spiti fauna he cites one species of Fenestella and one of Protoretipora from the lower horizon and no Bryozoa from the upper.

The record from China is equally meager. Kayser identified Synocladia sp., Polypora sp., Rhombopora lepidodendroides, and Fistulipora tuberosa from Lo Ping. From Kantschoufu Loczy notes Rhabdomeson cf. rhombiferum; from the Lantsankiang Valley Polypora fastuosa, Polypora sp., Septopora biserialis, Acanthocladia cf. anceps, and Callopora or Fistulipora sp.; and from Pupjao, in the province of Yunnan, Polypora koninckiana, Polypora cf. gingantea, and Fenestella or Polypora sp. These Chinese faunas certainly resemble the Guadalupian, in a general way, but are not sufficiently extensive to form a satisfactory index of relationship.

If we may assign Calamopora sp. to the corals, the only Bryozoa cited by Beyrich from Timor seem to be those which he identifies as Alveolites mackloti and Heliolites mülleri. The latter is presumably a Fistulipora, possibly of the type on which Dybowskiella was founded, and which is represented by F. grandis var. guadalupensis in our fauna. The Alveolites probably belong to the same genus, the vesicular tissue between the cells not being visible. It may even belong to the same species, or at least to a closely related one. In the apparent absence of vesicular interspaces this form somewhat suggests those which I have placed with Domopora, but the cells are not circular, and prominent star-shaped maculæ appear to be absent. Rothpletz, in his paper on the faunas of Timor and Rotti, cites several Bryozoa, namely, Fistulipora muller, Fistulipora? mackloti, Fenestella virgosa, and Polypora sp. The two former are the species which have already been mentioned in connection with Beyrich's report. Somewhat in contrast to Beyrich's figure, Rothpletz represents F. mülleri as having circular zoœcia. If he is exact in this particular the species certainly does not belong to the group of Dybowskiella, as I had supposed. On the other hand, as Rothpletz's figures represent the zoœcia not only as circular, but without a lunarium and with interrupted walls, there is some legitimate doubt as to whether the form is a Fistulipora at all. It rather suggests the Guadalupian form which I have referred to Actinotrypa. Rothpletz's Polypora has unusually coarse fenestration, and the branches instead of being persistent are represented as anastomosing. These characters suggest to me that the generic relations are really rather with Goniocladia or perhaps with Phyllopora.

The "Permo-Carboniferous" of Queensland and New Guinea contains rather numerous representatives of the Bryozoa. Etheridge distinguishes 1 species of Monticulipora, 4 species of Stenopora, 4 species of Fenestella, 1 species of Phyllopora, 3 species of Protoretipora, 1 species of Glauconome, 1 species of Rhombopora, and 1 species of Myriolithes. Aside from a few types of wide distribution and long range this fauna and that of the Guadalupe Mountains have very little resemblance, though it is difficult to tell much about the Australian fossils, owing to their poor condition and unsatisfactory illustration.

The Australian Bryozoa described from New South Wales by De Koninck comprise 1 species of Penniretipora, 1 species of Dendricopora, 6 species of Fenestella, 1 species of Protoretipora, 1 species of Retipora?, and 1 species of Polypora. All these seem to have come from the lower beds except a couple of Fenestellas and Protoretipora ampla, a fauna too meager to denote much relationship with the Guadalupian even if it existed. I have not referred any form to Protoretipora, but it is to be noted that De Koninck regards Polypora mexicana, which I have provisionally identified in the "dark limestone," as belonging to that genus.

Owing to the fact that most of the older writers and some of the more recent ones have failed to study their bryozoan faunas by means of thin sections, and that even among authors of the present day there is considerable variation of usage in the employment of generic terms, it is less possible with the bryozoans than with almost any other group to trust to mere lists unaccompanied by descriptions and especially by figures. On this account a good deal of the data contained in the Russian reports which I have consulted is rather unsatisfactory for the present purpose.

Stuckenberg's paper on the corals and Bryozoa of the upper mid-Russian Kohlenkalk is a notable exception, but in this case I find it almost impossible to distinguish the horizon of the different forms. This author cites only one species of Fistulipora, apparently of the general type of F. grandis var. guadalupensis. Fenestella is represented by 6 species, Polypora by 4, Carinella by 1, Penniretipora by 1, Coscinium by 1, Rhabdomeson by 1, Ascopora by 1, Orbipora by 2, and Archæopora by 1-20 in all. The generic representation is almost entirely different from the Guadalupian and need not be discussed. Aside from Fistulipora the only genera possessed in common are Fenestella and Polypora.

The position of many of these species seems to be in the Moskovian. Trautschold cites from this horizon Fenestella veneris, Polypora martins, P. irregularis, P. dendroides, Ascopora rhombifera, Ceriopora inaequabilis, Coscinium selliforme, and C. michelinia.

In his monograph on Gschelian Brachiopoda, Tschernyschew lists the following Bryozoa among the associated fauna: Dybowskiella, Geinitzella, Stenopora, Fenestella, Archimedes, Polypora, Coscinium, Thamniscus, Synocladia, Phyllopora, Archimedipora, and Penniretipora. The commonest to occur and the best represented in species appear to be Fenestella, Polypora, and Archimedes. The last genus, as I hardly need to mention, is found in eastern North America only in the Mississippian. In the western portion of the continent the horizon seems to be, in the few instances in which it has been found at all, in the upper Carboniferous, making this region in this particular, as in some others, comparable to the Russian section. This singular genus is of course not known in the Guadalupian nor in the underlying Hueco formation. In Utah, where I have found it in the upper Carboniferous, its associated fauna is such as to indicate a correlation with the Hueconian much more than with the Guadalupian series.

A good many genera are common to the Gschelian, as illustrated by Tschernyschew's list, and the Guadalupian, but they are such as are world-wide in distribution and very long in range, and do not necessarily indicate for the two faunas any very close relationship. Stuckenberg lists from this zone 15 species of Fenestella, 1 of Ptilopora, 12 of Polypora, 1 of Goniocladia, 2 of Penniretipora, 1 of Thamniscus, 1 of Synocladia, 1 of Ramipora, 2 of Dybowskiella, and 2 of Geinitzella. This list, as well as that of Tschernyschew, shows a considerable percentage of common genera. Perhaps the most significant of the genera which occur in both faunas is Goniocladia.

Passing over less copious notices of Gschelian Bryozoa, I find that Stuckenberg cites a somewhat less extensive list of the same group from the Artinsk, namely:

In the Kungurstufe the number is still less, consisting of only 3 species of Fenestella, 2 of Polypora, and 2 of Geinitzella. These lists indicate about the same community of generic types with the Guadalupian as those of Gschelian forms. The absence of Domopora from the Artinsk and of Geinitzella and Synocladia from the Guadalupian are deserving of remark, though there are a number of less important Guadalupian genera not found in the Russian beds.

Krotow's monograph on the fossils of the Artinskian sandstone contains reference to a number of bryozoan species, but unfortunately none of them is figured. He cites 6 species of Fenestella, 11 of Polypora, 2 of Phyllopora, 2 of Ptilopora, 2 of Penniretipora, 1 of Coscinium, 1 of Monticulipora, 1 of Stenopora, 1 of Rhombopora, and 5 of Vincularia. This list discloses a bryozoan fauna which is of a considerably different complexion from the Guadalupian.

From the Permian of Kostroma, Tschernyschew cites only Synocladia virgulacea, Fenestella retiformis, Stenopora columnaris, and Fistulipora lahuseni. The Permian Bryozoa cited by Netschajew consists of Fistulipora permiana (poorly figured but possibly not a Fistulipora at all), Geinitzella columnaris, and G. crassa together with 5 species of Fenestella, 9 species of Polypora, and 3 of Phyllopora. Golowkinsky cites Stenopora columnaris, Phyllopora sp., and Fenestella sp. In proportion as the bryozoan fauna of the Russian Permian has become reduced to a relatively few generic types of universal distribution it has lost character altogether. It lacks a number of genera, some of which are rather important in the Guadalupian, besides containing some which are non-Guadalupian, such as Batostomella and Synocladia. On the whole it can not be said that the faunas of the Permian of Russia and the Guadalupian show any marked relationship in point of the Bryozoa which they contain.

The only bryozoan cited by Abich from Djoulfa, in Armenia, is Polypora fastuosa, and by Enderle from Balia Maaden a species of Phyllopora and one of Fenestella.

I do not know whether Gemmellaro described the Bryozoa of the Sicilian fauna from Palermo, but if so, I have been unable to consult the work. Likewise, Schellwien's account of the Bryozoa of the Carnic Alps, if he prepared one, has escaped me. Angelis d'Ossat has published a report upon the corals and Bryozoa of the Carnic Alps, but the species while discussed are not figured. The following are recorded among the Bryozoa, constituting a fauna which has very little in common with the Guadalupian:

Gortani, who also fails to figure his forms, cites from this region a fauna closely related to that of Angelis d'Ossat:

From the Dyas of Germany Geinitz cites the following:

While most of these genera occur in the Guadalupian also, it is doubtful whether this fact should be regarded as very significant. The absence from the latter fauna of Synocladia and Batostomella (Stenopora columnaris), and the presence of Domopora are deserving of notice.

The Bryozoa of the English Permian are cited by King under the following titles:

This list, almost the counterpart of that of the Dyas, shows a moderately close relationship with the Guadalupian, but here again the absence from the latter of Synocladia, which seems to be something of a feature of the European Permian, is worthy of note.

From the south point of Spitzbergen, Toula cites Stenopora sp.; from Axel Island, Fenestella (1 species), Polypora (3 species), Ramipora (1 species), and Phyllopora (1 species); and from the cape between the arms of North Fjord, Stenopora ramosa, Stenopora tubulosa, Fenestella sp., and Polypora sp. Lundgren records only Stenopora columnaris. In these faunas from Spitzbergen I recognize no special affinity with the Guadalupian.

A rather extensive series of bryozoan forms is recorded by Toula from Nova Zembla, but this also seems but remotely related to the Guadalupian.

The occurrence of Archimedes in this Arctic fauna at an upper Carboniferous horizon is of considerable interest, connecting it with the occurrences in Russia and Utah.

From the West Sahara (Igidi) Stache cites only 2 species of Fenestella, 1 of Ascopora, and 1 of Stenopora?, and the associated fauna is probably much older than the Guadalupian.

Gabb cites Retipora from Peru and D'Orbigny Ceriopora ramosa and Retipora flexuosa from Bolivia.

After withdrawing species of western distribution, the Bryozoa of the typical American Pennsylvanian, according to Nickles and Bassler's catalogue, comprise 14 genera and 49 species, as follows:

Other species have since been added, but these do not materially affect the matter in hand. I may note, however, that Meekopora is now known in the Pennsylvanian as well as in the Mississippian.

Comparing this list with that of the Guadalupian we find that the two faunas have in common Fistulipora, Meekopora, Stenopora, Fenestella, Polypora, Acanthocladia, Thamniscus, Septopora, and Rhombopora, while the Guadalupian has Domopora, Leioclema, Goniocladia, Phyllopora?, and Actinotrypa?, not found in the Pennsylvanian, and the Pennsylvanian has Chainodictyon, Cystodictya, Diploporaria, Pinnnatopora, Prismopora, and Streblotrypa, not found in the Guadalupian. As to the genera which are present in both faunas, it is to be noted that in almost every instance the species representing them are different in each. Seldom, however, are the types so well marked and peculiar that the two faunas do not contain species more or less closely allied. In regard to the genera not held in common by the Guadalupian and Pennsylvanian faunas, the most significant on the part of the former are unquestionably Domopora? and Goniocladia, which lend the fauna a decidedly non-Pennsylvanian aspect. Actinotrypa and Leioclema, as is well known, are represented in the typical Mississippian, and the extension of their range into the Pennsylvanian may possibly be looked for, and would certainly be less of a novelty than the appearance of Domopora or Goniocladia.

Over half the Pennsylvanian genera have not been found in the Guadalupian, but in view of the still very partial knowledge which we possess of the latter fauna this number stands to be considerably diminished.

On the whole, I regard the bryozoan faunas of the Guadalupian and Pennsylvanian as rather closely related, probably more closely than the Bryozoa of the Guadalupian and any foreign fauna with which comparison has been made. They are, however, rendered very distinct both by the presence throughout of different species, though often of the same genus, and the presence in the Guadalupian of some novel and peculiar types, such as Domopora? and Goniocladia. I have found nothing comparable to the former genus in any of the Paleozoic faunas which have been consulted.

Family CERIOPORIDÆ Busk?

Genus DOMOPORA D'Orbigny?

In the Guadalupian series, especially in the lower beds of the Capitan formation, occurs an interesting and beautiful bryozoan type, which I have placed provisionally with D'Orbigny's genus Domopora. Although the position of the Carboniferous forms will probably prove to be in the taxonomic neighborhood of Domopora, it is not certain that they will find place directly with that genus. Nevertheless, I am not at present prepared to demonstrate their distinctness and to establish them as a new genus, for although the material examined is fairly plentiful, little of it is in a condition suitable for the study of zoarial structure. The majority of the specimens are silicified, and while it is thus possible by means of acid to obtain them free from inclosing rock, the processes of replacement have usually been such as to obscure or alter the details of structure in some degree. In specimens which have not been replaced by silica, moreover, the original test seems to be represented by dolomite, which has equally obscured the structure. Nevertheless, a certain amount of knowledge has been obtained of the microscopic structure of these forms from such examples, and occasionally from calcareous ones.

The zoœcia consist of cylindrical cells of nearly uniform size, which here and there at the surface tend to become confluent or to be connected by short grooves, especially in lines radiating from the maculæ. The zoœcia are interrupted by occasional tabulæ, though these structures are usually rare. They have no peristome nor any lunarium. Large maculæ having usually more or less of a stellate shape are a striking feature. At the surface the zoœcia are separated merely by the thickened walls, without either acanthopores or mesopores. In thin sections, so far as can be judged from our material, the thickened walls are not moniliform, as in Stenopora, but appear to be structureless or else with but small, dense granules, which may possibly represent obsolete acanthopores.

In some respects the form thus characterized may be compared to Batostomella, to Stenopora, and to Fistulipora, or, perhaps better, to Cyclotrypa. If in the latter genus the interspaces, instead of being occupied by mesopores, were filled with solid deposit, the structure would be like the forms under consideration. Similarly, that group of Stenopora which has thickened but not moniliform walls and few tabulæ, with sometimes very small if numerous acanthopores, seems but a step removed structurally from the group under discussion. Likewise, in Batostomella, if the transmutation of the mesopores into solid tissue were carried still farther, and followed by the gradual loss of acanthopores which replace them, one might imagine the present type to be related to Batostomella or even descended from it. A more complete knowledge of the minute structure of these Guadalupian forms, however, is necessary to such questions, as well as the discovery and investigation of intermediate types.

Although it is at present impossible to reach a satisfactory conclusion as to the relation of the present forms with D'Orbigny's genus, certain differences may yet be pointed out. Among the more obvious, the comparison being based on D'Orbigny's description and figures,a are the absence of series of large cells, often emerging along elevated rays (the present forms showing cells of but a single size), and the fact that colonies seem not to be formed by superimposed layers, though sometimes a few interruptions can be seen. Furthermore, in typical Domopora there appear to be no maculæ, at least of the type which is such a striking feature in the forms under consideration; but it is possible that Domopora should be reviewed on the basis of characteristic material from European beds.

---

aAnimaux invert., Terrain Cretace: Pal. française, vol. 5, Bryozoaires, 1850-1852, p. 986.

---

Domopora? shows a rather extensive specific development in the Guadalupian, as well as an abundance of individuals. The characters used for discrimination are less those of mensuration of parts than those of structure. In fact there often appears to be no more variation in the size of the zoœcia and their distance from one another in different species than in different individuals of the same species. The mode of growth, both of the zoarium as a whole and of its individual zoœcia, especially in relation to the macular areas, seems to form a practical and satisfactory method of discriminating species in this group.

This type of bryozoan, which appears to be rather Mesozoic than Paleozoic in its character, is not known in the Carboniferous except in this one area, faunas in other regions which are supposed to represent about the same horizon being, so far as known, without representatives of it. It therefore forms an individual feature of the Guadalupian fauna, and one which contributes considerable to its neo-Paleozoic facies.

DOMOPORA? TERMINALIS n. sp.

Pl. VII, figs. 19 to 21a; Pl. XVIII, figs. 1 to 6a.

The zoaria belonging to this species usually come in small subspherical masses, rounded above, somewhat prolonged and contracted below into a short, stout stalk.

The upper portion of the zoarium is poriferous; the lateral portions are closed. Such forms, however, though in our collection by far the most numerous, are perhaps only a partially mature condition, for some colonies have been found, with but little question developed from the type just referred to, which have an elongate or subcylindrical shape, produced by a considerably more extensive prolongation of the stalk. The small bulbous colonies are usually symmetrical, but they are sometimes rendered irregular by the character of the object to which they were affixed. Occasionally one side is developed more extensively than the others, as if the projections thus produced were the beginning of a bifurcation, but no really branched examples have been noted. The imperforate condition of the lateral areas is without much question a modification of a porous condition, the zoœcia having been closed over and apparently having received an additional testaceous coating. The lateral pores are less completely closed in the long cylindrical colonies than in the short spherical ones.

The terminal portion of the growing end in each zoarium is occupied by a macula, that is, an area upon which no cells are developed. The apertures extend in more or less regular rows from the borders of the macula, which projects rays of noncelliferous surface among them. The apertures are small, circular, separated by intervals of about their own diameter, and they come about eight in a space of 2 mm. in a radial direction. They vary somewhat in size in the same specimen and also in different specimens.

To a certain and somewhat limited extent the macula is a variant with the age of the colony, at least the large zoaria possess it in a high degree of development, while one or two specimens in which the feature is fairly obsolete are rather small. If it does vary appreciably with the age of the colony it at least is usually a recognizable feature in very young zoaria. In different specimens the maculæ also present certain variations in shape and in their relation to the rows of pores with which they interdigitate. This structure seems to be formed, in some cases at least, by an overarching platform, which conceals but does not entirely close the pores below; for in some examples it is possible from the side to look under this covering and see the cells, which are concealed when viewed from above. In one example from which the end has been broken, as would be expected, an earlier macula is exposed, apparently similar and subjacent to that which was doubtless terminal in the perfect condition.

The cells, as has been remarked, are circular in section without, so far as can be seen, any mesopores. By a thickening of the walls they are often separated at the surface at considerable intervals by a dense undifferentiated deposit. Underneath this outer flooring they are sometimes seen as tubes, which are not in contact; nor are the intervening intervals occupied in any manner except here and there by a small tube, which is probably a young cell. This may be the original condition, but of this I entertain some doubts, as the appearance might easily be due to the incomplete replacement to which these objects have been subjected. The interzoœcial spaces may have been solid below, as well as at the periphery, and all but the superficial silicified portions may have been removed by etching. The zoaria appear not to have been interrupted by tabulæ. They are nearly straight, with but a slight outward curvature, and slope strongly outward from an axial line.

As might be inferred, these colonies originate in an explanate zoarium having its external surface covered with an epitheca and cemented to some external object. When the object of attachment is large and regular a symmetrical colony usually results; where it is small and irregular an unsymmetrical colony; but irregularities also often occur from apparently intrinsic causes, such as the incipient branching, or what appears to be such, mentioned above. The basal portion is as a rule more or less expanded where attached, narrowing above and finally expanding to form the subspherical bulb bearing the zoœcia.

In point of size few colonies reach a diameter of 5 mm.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930), hill southwest of Guadalupe Point (station 2924), Guadalupe Point (stations 3762b, 3762d, 3762e), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 3500 and 2969).

DOMOPORA? OCELLATA n. sp.

Pl. VIII, figs. 7 and 7a; Pl. XVIII, figs. 7 to 10; Pl. XXVII, figs. 13 to 14a.

This species occurs in the form of elongate cylindrical stems, which occasionally seem to manifest a disposition to bifurcate, though no branching specimens have yet been found. The largest stems have a diameter of about 5 mm., and they occur as small as 3 mm. or less. This form, unlike Domopora? terminalis, does not have the lateral zoœcia closed, nor does it develop a single terminal macula. Instead, the entire surface, except possibly the point of attachment, has open zoœcia, among which, especially on the sides, are distributed large well-marked maculæ. The latter have a stellate shape and are characterized by the entire absence of zoœcia. They are arranged about 3-1/2 mm. apart from center to center.

The zoœcia are similar to those of Domopora? terminalis. They vary somewhat in size in the same colony, those nearest the maculæ being proportionately larger than the more distant ones. They are separated, as a rule, by intervals about equal to their own diameter, sometimes more and sometimes less. The interzoœcial integument appears to be solid, no traces of mesopores having been noted.

The internal structure, so far as it can be made out, without thin sections, from silicified specimens, is similar to that of D.? terminalis. The zoœcia bend outward, however, much more strongly, and a cross section of a branch shows two more or less distinct zones, an inner one, in which the tubes are transected more or less crosswise, and an outer one, where the course of the section is more or less longitudinal.

Well-preserved and characteristic specimens of this species are readily distinguished from Domopora? terminalis, but unfortunately many specimens are neither one nor the other. No especial profit would accrue from considering the obscurities which unquestionably arise from preservation, but it is sometimes not possible to discriminate with certainty between intrinsic and extrinsic characters. Our material seems to show that in the older portions of large colonies the zoœcia tend to close up through the thickening of the walls, thus solidifying the same and obscuring not only the arrangement of the zoœcia and maculæ, but their presence as well. In connection with the more terminal portion such occurrences would occasion little difficulty, but as all the large colonies are more or less fragmentary it is usually impossible to dispose of such examples satisfactorily.

Another difficulty arises from the opposite source. Very young colonies before they have become elongated stems, have very much the appearance of colonies of Domopora? terminalis of similar size. The lower portion is often similarly covered with a dense investment, and the length is scarcely sufficient to afford room for the lateral maculæ. The absence or presence of a terminal macula might serve as a practical criterion, but I am not sure that this is invariably present in an appreciable degree in D.? terminalis, and of course where only imperfectly developed it would so much the more easily be effaced by imperfect preservation.

One or two specimens represent another type which has been placed here with some hesitation. They are in the shape of irregular hemispherical masses having but a small diameter. They possess no distinct stalk or annular nonporiferous area, and the zoœcia are interspersed with large stellate maculæ, as in Domopora? ocellata. I have regarded these as representing a young stage of a rather large cylindrical colony. They might, however, be considered as belonging to an incrusting species.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966?); base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930), Guadalupe Point (stations 3762b, 3762d?, and 3762e), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2957, 2962, 2969, and 3500). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

DOMOPORA? CONSTRICTA n. sp.

Pl. XVIII, fig. 11.

This form, which is represented by several specimens, is related to Domopora? ocellata, but is distinguished chiefly by having the zoarium crossed by constrictions, which are located with reference to the maculæ, passing across their centers. In the type specimen the constrictions seem not to surround the cylindrical colony, but to be confined to the general region of macular development. Maculæ on the other side, not on the same circumferential line, seem not to be accompanied by constrictions. Aside from the feature just mentioned, there is little to distinguish this form from D.? ocellata, and I doubt whether it can be properly considered more than a variety of the other.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

DOMOPORA? VITTATA n. sp.

Pl. XVIII, figs. 12 and 12a; Pl. XXVII, figs. 15 and 15a.

Associated with Domopora? ocellata, and apparently derived from it, occurs a form which very much resembles D.? hillana. It attains this appearance by reason of the maculæ, which occur rather regularly alternating in two rows, taking on a lunate or fillet-shaped instead of a stellate configuration. The maculæ are usually elevated, as in D.? ocellata, but sometimes depressed. To make the resemblance to D.? hillana even greater the branches are sometimes more or less flattened or compressed. A difference which seems to me quite fundamental, however, is found in the method of cell construction. In D.? hillana the zoœcia spring from a median plate, while in the form under consideration they are directed radially. In addition to the structural difference the somewhat more compressed branches and slightly larger zoœcia of D.? hillana should serve as an additional means of distinguishing the two forms.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

DOMOPORA? HILLANA n. sp.

Pl. XXXI, fig. 19.

This handsome species is distinguished from its associates in the same genus by its growth more than by structural characters, the size of the cells, etc., being, so far as one can tell without thin sections, essentially as in Domopora? ocellata. Domopora? hillana, however, is a bifoliate form, which grows in rather slender, frequently bifurcating, flattened branches. So far as examined these seldom have a width of more than 2.5 mm., while the thickness is about 1.5 mm. This dimension diminishes toward the top of the branches. The zoœcia originate from a median plate, which is distinctly seen forming the long diameter of the branches, and extend somewhat obliquely forward. Large maculæ are developed along the edges of the fronds, extending onto both faces. Instead of being circular, they are fillet-shaped, the central and broader portion being situated on the edge of the frond and more proximal than the ends which point backward. The maculæ also as a rule form a depressed or constricted area, so that, quite additionally to branching, the edges of the frond present a notched or dentate outline.

This is a very pretty and strongly characteristic species. With good specimens it ought not to be possible to confuse it with any of the species yet discovered in the Guadalupian fauna.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

DOMOPORA? INCRUSTANS n. sp.

Pl. XXVII, figs. 16 to 16b.

The type and only specimen of this species formed a small elliptical expansion bent along its longest diameter around some object which has now disappeared and to which it was attached, so that at present it possesses an annular shape. Its inner and lateral surfaces are covered by an epitheca, and its outer surface is poriferous. The zoarium is about 1-1/2 mm. thick in the thickest portion.

The size of the apertures in general appearance is similar to that seen in Domopora? ocellata, but there are no stellate maculæ. This feature is represented by non-poriferous rays diverging from certain points on the margin of the zoarium, as if from maculæ whose centers lay without the present complete limits of the zoarium.

The relations of this form are doubtful. It is possibly but a very aberrant initial colony of D.? ocellata, but it differs so much from such mature or even youthful colonies as are known to me (or of D.? terminalis as well) that it did not seem justifiable, without further evidence, to refer both to one species.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Family FISTULIPORIDÆ Ulrich.

Genus FISTULIPORA McCoy.

FISTULIPORA GRANDIS var. GUADALUPENSIS n. var.

Pl. XVII, fig. 18; Pl. XXV, fig. 7; Pl. XXVII, fig. 17.

This form usually occurs as cylindrical masses of various dimensions, which sometimes attain a diameter of 20 mm. but are often smaller. The longest fragment yet noted is about 50 mm. in length. It is evident that these vertically elongated masses must have arisen from an initial flat expansion, and in fact specimens having this shape are also found possessing apparently all the characters of the cylindrical examples. Some of these explanate specimens have a greater diameter than any of those which are cylindrical, and it seems not unlikely that the latter may have been attached by expanded bases, or that the shape may have varied from cylindrical elevations to rather thick flat expansions having nodular areas rising above the main mass. Cylindrical specimens frequently have the lateral surface more or less completely covered with an exterior investment resembling an epitheca. In places this is sufficiently heavy to completely mask the cells beneath. In places also the cells are partially visible, while elsewhere they appear to have been open and functional.

In transverse sections the zoarium is seen to consist of the usual autopores and mesopores, in the arrangement of which there is the greatest latitude. The zoœcia vary somewhat in size, but usually are about two-fifths of a millimeter in diameter. They are characterized by a very strongly marked lunarium, which surrounds about half of the zoœcial tube, or a little more. The zoœcia usually run about four to four and a half in a distance of 2 mm., but may occur in as great number as five and a half. In macular areas, however, they are of course farther apart, and in these regions spaces of 2 mm. can sometimes be measured in which none at all are found. The mesopores vary greatly in size. Occasionally they reach as large dimensions as the zoœcia, but usually they are about half that size. They occur as a rule in one or two rows between the zoœcia, but, as already remarked, in maculæ they are much more abundant. Usually the zoœcia are situated from one-half to a full diameter apart, but sometimes they are almost in contact.

In longitudinal section corresponding variations are shown. The mesopores exhibit a wide range, not only in the number which intervene between adjacent zoœcia, but also in the size and frequency of tabulation. From 10 to 24 tabulæ occur in 2 mm., and the structures may make segments which are twice as wide as long; or, on the other hand, twice as long as wide. The zoœcia are rather sparsely and irregularly tabulate. In places; especially near the surface, they are often uninterrupted for as much as 2 mm. or more, but below, three or four tabulæ may occur at distances of one-fourth of a millimeter apart or less.

This species is readily distinguishable from the two known species of the American Pennsylvanian. From F. carbonaria it differs in its mode of growth, in its more abundant mesopores, and in its larger and more highly developed lunarium. From F. nodulifera it differs in its mode of growth, its larger cells, and its more extended lunarium.

Several species have been described from India which are also related to this. The growth is similar to F. grandis. Minor differences connected with the size of the zoarium, of the individual cells, of the abundance of maculæ, etc., might be pointed out, but the most striking distinction is to be found in the much larger and more embracing size of the lunarium in the American form. About the same difference exists between the latter and F. expansa, and there is the additional one of mode of growth, F. expansa, as the name suggests, forming flattened expansions. F. parasitica Waagen and Wentzel forms incrusting lamellæ similar to the American F. nodulifera and has the lunarium less well developed.

Waagen and Wentzel included the first two species in their genus Dybowskiella. When that name was first introduced the genus Fistulipora was imperfectly understood, and Dybowskiella would seem to have been well established; but at present the weightiest authorities seem to regard Dybowskiella a synonym of McCoy's genus, a conclusion which my own judgment entirely sanctions.

F. grandis var. guadalupensis is, with possibly one exception, the commonest bryozoan of the Guadalupian fauna, having been recognized at a large number of localities representing the horizons of the Delaware Mountain sandstone and the Capitan limestone, both Shumard's "dark limestone" and his white. A good many of these identifications, however, have been made on silicified material, which is not only refractory in making sections but has also been more or less altered in appearance in the processes of replacement. Where the silicification is good the lunarium projects as a strongly elevated spine, which on examination is seen to have a crescentic shape. Sometimes the cells appear to be more or less completely closed over in a sort of pustular elevation. This last may in fact truthfully represent the original calcareous condition, but there are various other aspects which are more clearly accidental to fossilization.

Horizon and locality.—Top of Capitan formation (stations 2966 and 3762a) and middle of Capitan formation (station 2926), Capitan Peak; "dark limestone," Pine Spring (station 2930), hill southwest of Guadalupe Point (station 2924), Guadalupe Point (stations 3762b, 3762e, 3762d); Delaware Mountain formation, Guadalupe Point (stations 2919 and 2963), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2957, 2962, 2969, and 3500). Delaware Mountain formation, Comanche Canyon, Glass Mountains (station 3763), and mountains northwest of Marathon (station 3840), Texas.

FISTULIPORA GUADALUPAE n. sp.

Pl. VIII, figs. 5 and 5a.

This species grows in thin incrusting expansions of undetermined extent but very small height. In the specimen studied the latter dimension is less than 2 mm.

The zoœcia are very various in size, the average among the smaller ones being only about 0.1 mm., while they range up to 0.2 mm.a They are also variously disposed, being sometimes 0.1 mm. or less apart and at others probably as much as 0.3 mm. There seems to be a certain compensation in these arrangements, the larger cells being close together and the small ones far apart. The cross section of the zoœcia is nearly circular. A lunarium is fairly well developed, but does not materially indent the outline. It embraces less than half of the zoœcium. The spaces between the zoœcia are occupied by two or more series of rather small vesicular mesopores. Tabulæ seem to be wanting in this form as seen in longitudinal sections. The external surface is strongly undulating, by reason of numerous closely set, rather strong monticules.

---

This figure is drawn from a portion of the section where the zoœcia are of the smaller size and nearly uniform.

---

This species is distinguished by its incrusting lamellar habit of growth, very small zoœcia, and prominent, closely arranged monticules, characters which, so far as known, are not found combined in any other American species.

Horizon and locality.—Top of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2966).

FISTULIPORA sp.

This form is represented by a small silicified specimen which permits the determination of only a few of its distinctive characters. The growth is in highly contorted laminæ a little over 2 mm. in thickness.

This species is at once distinguished from F. grandis var. guadalupensis by the small size of its zoœcia, which are separated by distances equal to one or two zoœcial diameters. About six occur in a distances of 2 mm., but the conditions are such that an accurate count is difficult. So far as can be made out the lunarium also, which is so strongly developed in the other species, is here inconspicuous. Further particulars are not furnished by the material in hand.

Horizon and locality.—Delaware Mountain formation. Comanche Canyon, Glass Mountains, Texas (station 3763).

Genus MEEKOPORA Ulrich.

MEEKOPORA sp.

Pl. XXXI, figs. 18 to 18b.

This species is represented by a single specimen, which is fragmentary and silicified. Thin sections have not been made, and some of the characters have not been determined.

Owing to the fragmentary condition of the specimen the shape and general nature of the frond can not be described. It appears to have been at least stoutly constructed, as the specimen has a diameter of a little over 4 mm.

The zoœcia are distributed rather evenly on the surface and quite up to the edges. Maculæ are small and not very conspicuous. The zoœcia vary considerably in size and shape, and while in a general way their distribution is regular they also vary in the distance which separates them. Four apertures and three interspaces, or sometimes four apertures and four interspaces, occur in 2 mm. in nonmacular areas. They usually stand about their own diameter apart, but may be confluent or separated by one and one-half times their own diameter. They average about 0.3 to 0.4 mm. in diameter, as well as can be determined. In the present specimen the apertures possess no peristome, and it is doubtful whether there was an appreciable lunarium, though this is by no means certain. The number of mesopores is likewise uncertain, but there appear to have been three or four rows between the zoœcia. The zoœcial tubes slope gently but distinctly in both a distal and a lateral direction. They appear to have had a few tabulæ situated more toward the base of the tubes than near their mouths. Probably not more than one or two were developed in each zoœcium.

This is clearly distinct from Meekopora prosseri, the only "Coal Measures" species as yet described from American rocks. It is also with but little question distinct from the lower Carboniferous species of the genus, though the differences, which are quite obvious on comparison, need not be recited here. The Guadalupian form therefore appears, as might have been expected, to be an undescribed species, but it has not seemed desirable to distinguish it as yet by a new name, because its characters are hardly known with sufficient fullness and certainty to discriminate it from more closely related forms whose discovery is probable.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Family BATOSTOMELLIDÆ Ulrich.

Genus STENOPORA Lonsdale.

STENOPORA GRANULOSA n. sp.

Pl. XIX, figs. 1 to 1c.

This species grows in thin lamellar expansions consisting of one or two layers of zoœcia, each of which has a thickness of about 0.5 mm. or less. The zoœcia rise with strong obliquity to the lower surface, and have occasional or moderately frequent mesopores. As a rule they are separated by rather thick walls, which have on an average a diameter of one-third to one-fourth that of the zoœcia themselves. The outline of the latter in transverse section is usually curvilinear, the thickness of the walls taking up most of the difference between the regularly curved and the polygonal figure. Acanthopores are fairly numerous, and all are of the granular type, none of the normal concentrically banded kind having been observed. Usually they are scattered and situated in the triangular thickening of the wall where three cells corner, or side by side where a thick wall separates two adjacent zoœcia, but occasionally they form relatively broad continuous bands through the center of the thick walls, the individual acanthopores being indistinguishable. Tabulæ seem to be completely wanting. The zoœcia are not readily followed in straight lines, but seven, or sometimes eight, come within a linear distance of 2 mm.

This species is especially characterized by its delicate lamellar growth, by the absence of diaphragms, and by the peculiar structure of the walls. It belongs to a section of the genus, well developed in Pennsylvanian time, of which Stenopora? signata is a representative form. It is of course very distinct from S.? signata, having a different mode of growth, more numerous mesopores, no acanthopores of the normal type, and differently arranged ones of the granular type, since those of S.? signata are more regularly distributed, are usually more numerous, and seem never to lose their identity in forming continuous granular bands, even when most closely arranged.

In its mode of growth as well as other characters this seems to be distinct from any of the species recognized in the Salt Range of India. It is true, however, that Waagen and Wentzel do not give very complete descriptions of these species, not mentioning, for example, whether the acanthopores are granular or concentric, etc.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

STENOPORA GRANULOSA n. sp. ?

Pl. XXIV, figs. 3 to 3c.

The form for which this identification is employed is known from very scanty material. In its mode of growth and structure it comes in many respects close to Stenopora granulosa. Some differences of moment can be pointed out, but as it is not clear that they may not be due to different stages of growth it seems best not to introduce for it a new name, on the one hand, or to confuse it with typical S. granulosa, on the other.

The mode of growth is essentially the same as the species with which it is provisionally identified, but the cells are somewhat smaller, nine, or even ten, occurring in 2 mm. in a not entirely straight line. The cell walls are considerably thicker averaging about one-half of a zoœcial diameter, or even more. Mesopores seem to be entirely lacking. A few acanthopores of the concentrically constructed type are scattered here and there. A few granular acanthopores seem to be equally disseminated, while the median portion of the thick walls is rather profusely granular.

The somewhat finer construction of this form may possibly indicate a specific distinction, but the thick walls and consequently smaller zoœcia, the absence of mesopores, and the presence of concentric acanthopores, with a slightly different construction of the walls themselves, are differences very marked in the two specimens studied, though they may be the result of difference in maturity. This form is still more like Stenopora? signata than is typical S. granulosa, but differs in the mode of growth, smaller zoœcia, thicker walls, less numerous and smaller normal acanthopores, and the more generally granular condition of the walls as distinguished from the individual granular acanthopores of the Pennsylvanian species.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2967).

STENOPORA POLYSPINOSA var. RICHARDSONI n. var.

Pl. VIII, figs. 6 to 6b.

This species forms cylindrical branching (?) zoaria, that of the type specimen having a diameter of 3 mm. The maximum diameter noticed is 5 mm. The macroscopic characters are unknown. The zoœcia are subangularly polygonal in outline, and separated by walls from one-fourth to one-half their own diameter. Six zoœcia occur in a linear distance of 2 mm. Mesopores are very rare. Acanthopores are numerous, small, and apparently of but a single size, though a certain amount of variation is shown in this particular. Some doubtful instances, however, of acanthopores two or three times as large as the usual size have been noted. The distribution of the acanthopores is irregular. In many specimens they are about their own diameter apart and occur in single rows, but in others their distance is somewhat less or greater, sometimes much greater, while occasionally a considerable space is left without any at all. Again, where the wall is especially thick, they are fairly numerous, but so irregularly disposed that no expression can be formulated as to how many rows they constitute.

In longitudinal section the walls are seen to be somewhat thickened in the mature region, but the characteristic nodose structure is wanting. Small points of greater density, whose arrangement is more or less regular, are a striking feature of the mature region, a character which has been noted also in Stenopora signata. Tabulæ are very scantly developed, practically absent.

This species is related to Stenopora signata, but is clearly distinct, having considerably larger zoœcia and smaller and more numerous acanthopores, which if not altogether of one size are yet very nearly so, the large ones being extremely rare. It is also related to Stenopora spissa, a significant and well-marked difference being found, however, in the thicker walls and larger acanthopores of Rogers's species. Much more closely related, however, is Stenopora polyspinosa, described by Condra, and I am uncertain how far the differences shown by the Guadalupian species, which is known from very scanty material, would be increased or diminished by more complete information regarding it, and by actual comparison of specimens. In view of the widely different fauna with which it is associated I feel disposed to give due weight to such as have been noted. It appears to have had less numerous, somewhat more widely separated and irregularly distributed acanthopores, of which there are fewer of the large size, and perhaps also fewer mesopores, which Condra describes as being one-fourth as numerous as the zoœcia. At present these differences are real enough, though not very great in degree; otherwise the two species, so far as known, possess almost identical characters.

A strongly developed specimen from station 2969 has the walls in the mature region considerably thickened so that they average one-third to one-half the diameter of the zoœcia, which have quite lost their polygonal outline. The other characters are the same.

A silicified specimen from station 3500 has been placed here with doubt. In the main it possesses the characters outlined above, differing only in having tabulæ which if not numerous are relatively so.

Horizon and locality.—Top of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2966). Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2963?, 2969, 3500?).

STENOPORA sp.

This form was found as an incrustation upon a small cylindrical or spinelike object, and apparently formed a zoarium by means of thin superimposed layers.

The typical specimen, however, consists of only one such layer having a maximum thickness of not over 1 mm. The macroscopic characters are unknown. In tangential section the zoœcia are seen to be small and circular, with thick walls. There are usually about ten zoœcia in the space of 2 mm., and the interval between two adjacent ones is usually about one-half a zoœcial diameter, though sometimes considerably less. Mesopores are practically absent. Large acanthopores are situated in the angles between three zoœcia. Usually along the mesial line of the walls connecting the acanthopores is a row of granules, which may represent small acanthopores, though they are of small size, irregular, and indistinct.

In sections passing longitudinally through the zoœcia these are seen to be short, having the walls highly thickened but not moniliform from the point where they bend into an upright position. Tabulæ appear to be practically absent.

So small a fragment was obtained of this species that it is feared it does not show the mature and characteristic condition of the colony. It appears, however, to be an undescribed form, and is distinguished by its mode of growth, very small zoœcia, thick walls, absence of mesopores, etc.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

Genus LEIOCLEMA Ulrich.

LEIOCLEMA SHUMARDI n. sp.

Pl. XIX, figs. 2 to 2d.

This species grows in cylindrical, sometimes bifurcating branches, which seldom exceed 3.5 mm. in diameter, and are often smaller.

The apertures show no serial arrangement in any direction, but appear in general to be rather regularly distributed. Nevertheless, when examined critically they are seen to vary in their relations to one another. Usually the distance from one zoœcium to those which are nearest is from one and one-half to two zoœcial diameters, but it is not difficult to find linear intervals which are as much as three or four times the ordinary. These interspaces are occupied by mesopores, of which there are usually three or four rows between adjacent zoœcia, and by acanthopores. The mesopores vary much in size, some being as large as the zoœcia, but as a rule they have only one-half or three-fourths the diameter of the latter. The average diameter of the zoœcia is from 0.15 to 0.20 mm. About five zoœcia and five interspaces occur within a distance of 2 mm.

The zoœcia are rather closely tabulate, though the tabulæ are usually separated by intervals greater than a zoœcial diameter. The mesopores are also abundantly and irregularly tabulate, the distance between these structures being sometimes greater and sometimes less than the diameter of the mesopore.

This form is well distinguished from the several Mississippian species of the genus, no species of Pennsylvanian age having thus far been described from American rocks. The most closely related species is unquestionably L. punctatum of the Keokuk, a comparison with which shows that in L. shumardi the zoœcia, while of about the same size, are in the mature region placed at wider intervals—that is, there are more rows of mesopores between them. The acanthopores are at the same time less numerous and of distinctly larger size. The tabulation of the zoœcia is also more abundant.

Leioclema shumardi is perhaps more nearly related to the species from India and Europe which Waagen and Wentzel figure as Geinitzella columnaris, especially that form which they designate as var. ramosa multigemmata. From this, however, the Guadalupian species is distinguished by the large size and great development of the mesopores and by the wide spaces which separate the zoœcia. The authors mentioned above give no direct measurements of the species in question, but from their figures and the magnification which they are said to show there are many more zoœcia in a given distance, in fact well-nigh twice as many.

These authors place in the synonymy of Geinitzella columnaris Schlotheim. Geinitz's species Stenopora mackrothi, which under the name of Chætaetes mackrothi Shumard cites from the "dark limestone" of the Guadalupe Mountains. It seems probable that the form thus referred to by Shumard is the one under present consideration.

The typical specimens of this species were obtained in the "dark limestone." From the white limestone of the Capitan formation a few examples of Leioclema have come to hand which present in the main the same characters as L. shumardi and yet show certain differences of apparently minor importance. For the present, therefore, these later representatives are placed in the same species with those which they succeed. When both are better known their relations can be better determined.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966?); middle of Capitan formation, Capitan Peak, (station 2926?); base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930), and Guadalupe Point (stations 3762d?, 3762e), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969?).

Family FENESTELLIDÆ King.

Genus FENESTELLA Lonsdale.

In our collections the genus Fenestella forms but an insignificant factor in the Guadalupian fauna, although when the material was gathered this class of fossils was by no means neglected. It is true that thirteen types have been discriminated, indicating that the genus was present in variety if not in abundance, and it is even possible that at favorable localities and horizons abundant material also could be obtained.

Most of the specimens at present known from this horizon are in a silicified condition and were obtained by etching. They are mostly small fragments, and in general not well preserved. Under ordinary circumstances it would have been permissible to pass over with some general mention a part of the fauna in such condition, but so much adventitious interest attaches to even this small and imperfect element, by reason of the novel character of the whole, that it appeared to me desirable to afford it more careful attention. On the other hand, the data obtainable in many of the forms was too incomplete to permit satisfactory treatment from a specific standpoint. Thus, while it has been possible to identify only one of these varieties with species which have already been described, I have felt justified in describing only two or three as new, an unfortunately large number being merely described, so far as description was possible, in an anonymous manner.

It might be inferred from the very individual character of the associated fauna that the Guadalupian Fenestellas would present altogether novel types when compared with those found with Pennsylvanian faunas of the Mississippi Valley. This has not proved to be the case. While it may be said with some confidence that nearly if not quite all the Guadalupian species are different, they present no especially striking developments nor, so far as I have been able to ascertain, any marked individuality considered as a whole. The only significant fact which my study of these forms seems to develop is that, like the rest of the fauna, they are different from Pennsylvanian forms of the interior region, though much more analogous than is the case with the brachiopods. As a rule the Guadalupian species are rather fine regular forms, often delicate but sometimes more solid.

The exploration of Shumard in this general region brought to light, as is well known, a small number of forms which were described by Prout nearly fifty years ago. Only one of these was obtained at the locality and from the horizon at which my collections were made, and it is unfortunately not included with them, Prout's description being neither very complete nor accompanied by figures. As will be expected, most of the species described by him from this region, since they were collected at much lower horizons than the Guadalupian, prove to be distinct from anything in this report. As a rule they are larger and more robust types, and will doubtless be found in the Hueconian fauna (to which horizon they belong), which I hope later to describe.

Waagen discriminated only two species of Fenestella in the Salt Range, from which it would appear that the genus is even less well represented than in the American fauna. Neither of Waagen's two species appears to be identical with those from the Guadalupe Mountains, which really find closer analogies in the Pennsylvanian rocks of the Mississippi Valley, a circumstance not altogether surprising in view of the much greater number and variety of species found there. It may be mentioned that one of the two species discriminated by Waagen is identified by him as F. elegans, which Meek described from Nebraska. The identification appears, to be very close.

FENESTELLA POPEANA Prout.

1858. Fenestella Popeana. Prout, Trans. Acad. Sci. St. Louis, vol. 1, p. 229 (date of volume, 1860).

White Permian limestone: Guadalupe Mountains, New Mexico.

1859. Fenestella Popeana. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 388 (date of volume, 1860).

White and dark [Permian] limestone:. Guadalupe Mountains.

Corallum most probably campanulate, rapidly curving outward from frequent bifurcation. Longitudinal rays or interstices subangular, striated as seen by the impression on the cortical envelope of the reverse; keel obsolete; bifurcations frequent, mostly about one line apart, large near the base, nearly as wide as the fenestrules. Dissepiments moderately large, round, expanded at junction with interstices. Fenestrules ovate or quadrangular, rounded at the angles, five in the space of two lines longitudinally, about five transversely. Cells large, ovate, directed upward and outward to the axis of interstices, alternate on the two sides of the longitudinal ray, three to each fenestrule, rarely four, caused by a supernumerary placed at the angle of bifurcation.

This beautiful species is dedicated to Capt. John Pope, whose indefatigable labors in the service of his country and whose zeal and devotion to the interests of science deserve the compliment. It was collected, with other specimens, from the Guadalupe Mountain, by Dr. George G. Shumard, and is classed by our worthy president, Dr. B. F. Shumard, as a Permian species. The description is drawn from a fragment of one side of the expansion; but its form, we think, can be inferred as campanulate. Only a small portion of the poriferous side is preserved, the fracture being mostly down to the cortical portion of the reverse; sufficient, however, can be made out to identify it as a new species.

Comparisons.—It resembles very nearly F. patula (McCoy), but the latter has larger interstices, with a strongly marked keel. It is only half as large as the F. Popeana, and, besides, the latter has fenestrules nearly double as wide as the interstices, being at the same time strongly corticated, at least on the reverse.

It resembles the F. antiqua (Gold sp. McCoy), but differs by the thickness of its interstices, as well as by the greater length and fewer number of fenestrules in a given space.

Locality.—Permian white limestone, Guadalupe Mountain, New Mexico.

The foregoing is Prout's original description of this form, which seems to be distinct from the single species in our collection from the same locality and horizon and has not been recognized among the fragments thus far obtained from other localities and horizons in the Guadalupian.

FENESTELLA HILLI n. sp.

Pl. XIX, figs. 3 and 3a.

This species consists of a fine regular mesh composed of straight branches and stout dissepiments, which are not depressed below the nonporiferous surface. The fenestrules are about twice as long as wide, and vary in shape from subelliptical to subquadrate. There are about four rows (and five branches) and longitudinally four fenestrules (and three dissepiments) in a distance of 3 mm.

On the nonporiferous side the branches seem to be smooth. On the obverse they are traversed mesially by a high carina, with an expanded top, along whose center is a row of small perforations, apparently about three opposite each fenestrule. The zoœcial apertures are not well shown, but they appear to be arranged one opposite each dissepiment, with two intermediate. A variation from this arrangement places three of them opposite each fenestrule, without any opposite the dissepiments.

Of species found in the Mississippi Valley this form is perhaps most closely similar to Fenestella wortheni, from which it may be distinguished by its more robust proportions. F. hexagonalis also is related, but clearly distinct. Under the name of F. corticata Prout has described a form somewhat similar to that under discussion, but having more zoœcial apertures to the fenestrule, and apparently a row of nodes down the back of each branch.

Horizon and locality.—"Dark limestone," Guadalupe Mountains, Texas (station 3762e).

FENESTELLA CAPITANENSIS n. sp.

Pl. VIII, figs. 4 and 4a.

This species includes a single specimen from the white limestone (Capitan), of which the nonporiferous face is exposed to view.

The zoarium had a shape more or less approximating to that of a portion of an inverted cone. The length of the frond as represented by the specimen is 17 mm. and the width about the same.

Branches and dissepiments, while the latter are distinctly smaller, are nearly of one size. The dissepiments expand at their junction with the branches, giving the fenestrules a rounded outline, which rarely becomes noticeably quadrate. The width of the dissepiments, while sometimes nearly equal to the length, is as a rule distinctly less. There are five rows (and six branches) in the space of 3 mm., and longitudinally four fenestrules (and five dissepiments) in the same distance.

The surface of the branches is smooth on the nonporiferous side, except that a distinct tubercle seems to have been developed where the dissepiments and branches intersect. The poriferous side, as already remarked, is not well shown. The arrangement of the zoœcia appears to vary considerably. Sometimes they are placed one opposite each dissepiment, with another intermediate, and sometimes there are as many as three intermediate ones, with all gradations between.

A common arrangement is for three apertures to open against a fenestrule without any opposite the dissepiments. As already intimated, the characters of the celluliferous face are largely obscured. A carina, if present, was probably low. At the same time there is evidence of a row of distant spines dividing the two series of pores. These statements, however, are tentative, needing future confirmation.

This species is very near the proportions of the form designated Fenestella sp. a, but is somewhat more massive in its construction, with more rounded fenestrules, and with the nonporiferous face tuberculate instead of smooth. What additional differences would appear if both forms were completely known can not be conjectured.

While comparable in some respects to several forms found in the Pennsylvanian of the Mississippi Valley, this species is distinguished from most of those which it resembles in other respects by its solid, compact structure, resulting from the relatively thick dissepiments, and by the presence of nodes on the reverse face. Its nearest allies seem to be F. conradi and the variety compactilis. It is a little more robust than F. conradi, with somewhat thinner dissepiments and with more elongate fenestrules, which have more numerous apertures opposite them. The reverse face of F. conradi is also smooth.

It is evidently distinct from F. popeana, as represented in Prout's description, a species which apparently was found at about the same locality and horizon. The sculpture is very different and the proportions somewhat so.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2905), and middle of Capitan formation, Capitan Peak (station 2926), Guadalupe Mountains, Texas.

FENESTELLA GUADALUPENSIS n. sp.

Pl. XIX, fig. 5.

This form consists of a rather coarse, somewhat irregular network. The branches are straight, slender, round, and separated by nearly equal intervals of two to three times the width of their own diameter. The dissepiments are slender, round, and distinctly smaller than the branches, but approximately equal to them. They are depressed and irregularly disposed, both as to distance and direction. The fenestrules consequently are of different sizes and shapes, occasionally wider than long, usually longer than wide, but always distinctly quadrangular in outline. There are three rows of fenestrules and four branches in a distance of 3 mm., and two or three fenestrules (and four dissepiments) or less in the same distance longitudinally.

The nonporiferous side is smooth. On the poriferous side the branches had a median angulation, which may have been a distinct through certainly not a high keel. The apertures are small and give rise to a wavy outline on the margins of the branches. There is usually one aperture opposite each dissepiment, and intervening between these from two to four, depending somewhat on the length of the fenestrule.

Of species found in the Mississippi Valley this form is clearly most nearly related to Fenestella delicatula. The description of that species would indicate that the two forms were similarly constructed, but that from the Guadalupe Mountains when compared with a somewhat fragmentary example from Seville, Ill., proves to be more robust in its growth, the dissepiments especially being relatively stronger. The surface of F. delicatula, furthermore, is ornamented by fine granulose liræ, no trace of which is found upon the form under discussion.

Three species from the same general region, though from a lower horizon, demand comparison. I refer to F. albuquerqueana, F. intermedia, and F. variabilis the first-named species, recently described by Bendrat, being probably a synonym for the older one founded by Prout (F. variabilis). These two, with their striated surface and prominent carina, are safely distinct from the present species. While I have little doubt that the latter is distinct from F. intermedia also, it is not easy, without specimens of that species, to point out differences which should distinguish them. It would appear, however, that there were fewer fenestrules to a given distance longitudinally in Prout's species.

It might be expected that this form would be the same as F. popeana, described from essentially the same locality and horizon, but while showing characters in common it is unlikely that the two species will prove to be the same. That described by Prout would appear to be more regular, with fewer fenestrules (longitudinally) in a given distance and fewer zoœcial apertures to a fenestrule.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

FENESTELLA GUADALUPENSIS var.

This species is represented by a single specimen in a thin section. Three rows and three branches occur in a distance of 3 mm. transversely, and two and one half fenestrules and two dissepiments in the same distance longitudinally. The branches are about twice as wide as the dissepiments and about two-thirds as wide as the fenestrules. The latter are subquadrate and about twice as wide as long. Apparently three cells lie opposite each fenestrule. In addition to the usual finely porous character of the test the noncelluliferous side of the present species was pierced by rather large-sized pores, in a general way so distributed that one is opposite to a dissepiment and one intermediate.

In its measurements this form is in close agreement with Fenestella guadalupensis. The chief differences recognized consist in the relatively narrow branches, while the large-sized pores, if, as seems not unlikely, they are the result of nodes or spines upon the exterior, clearly discriminate the present form. If the pores are due to nodes or spines, this form recalls F. capitanensis and Fenestella sp. f in this particular more than any other Guadalupian species, but the proportions are very different.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

FENESTELLA SPINULOSA Condra ?

Pl. XIX, figs. 4 and 4a.

1902. Fenestella spinulosa. Condra, Am. Geologist, vol. 30, p. 343, pl. 21, figs. 4, 5.

"Coal Measures:" Roca and Dawson, Nebr.

1903. Fenestella spinulosa. Condra, Nebraska Geol. Survey, vol. 2, Pt. 1, P. 55, Pl. 10, figs. 1-5.

"Coal Measures:" Roca and Dawson, Nebr.

This species consists of a regular, fine network of straight branches and dissepiments. There are five rows of fenestrules (and six branches) and longitudinally five fenestrules (and six dissepiments) in the space of 3 mm., or possibly a little less in both cases.

The dissepiments are considerably slenderer than the branches and somewhat depressed below the level preserved by their upper surfaces. The fenestrules are usually a little longer than wide, their outline as a rule being more or less distinctly quadrate.

The outer surface of the branches appears to be smooth. The inner surface is raised into a rather thin, high, median carina, which seems to develop a row of median nodes disposed in much the same manner as the zoœcial apertures. The latter are usually so arranged that one occurs opposite each dissepiment, and one halfway between; but occasionally an aperture occurs on either side of the dissepiment without an intermediate one.

Such characters are retained by the small but rather well-preserved fragment. Another small example has been referred to the same species, though it departs in certain details from that on which the description depends. The proportions are almost the same, but the construction is distinctly heavier; the branches and dissepiments are thicker, and the fenestrules are less quadrate. The poriferous side of this example is imperfect. It appears to have had two zoœcial openings opposite each fenestrule, but the arrangement may prove to be as in the first specimen.

Here has also been referred a specimen from the Capitan formation (station 2966), which is embedded in limestone and shows only the obverse face. It is somewhat exfoliated and obscured. The dimensions are essentially the same as those of the specimen from which the first description was drawn. The dissepiments appear to be a little larger, but this may easily be due to their difference in preservation. The apertures are as a rule so arranged that two of them are opposite each fenestrule. Occasionally there is also one opposite a dissepiment, but such cases are single, none having been observed where they regularly occur two opposite each fenestrule and one opposite each dissepiment.

In some of its characters this species can be compared with a number of Pennsylvanian forms from the Mississippi Valley, but it appears most to resemble Fenestella parvipora, F. perelegans, and F. spinulosa. The first-mentioned species hardly demands careful comparison, and the second, which is imperfectly known, is said to have the carina nearly obsolete. With F. spinulosa, however, as described and figured by Condra, the agreement is very close. Perhaps the only difference at present available is that the Guadalupian form seems to lack the faint striæ which F. spinulosa is described as possessing. So different, however, are the faunal associations of the former that it seems little likely to prove identical with Condra's species could well-preserved examples of both be brought into comparison. Fenestella shumardi, a species described from the same general region as the one under consideration, but from a different horizon, shows many points of similarity, but it can safely be concluded that this is not the species which Prout was describing.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969?).

FENESTELLA TEXANA n. sp.

Pl. XXVIII, figs. 9 and 9a.

A rather small fragment is all that remains of this species; it shows the following characters:

The growth is fairly even and regular. About three and one-half rows of fenestrules (including four branches) occur in a distance of 3 mm., and about the same number in the same distance in a longitudinal direction, including also four dissepiments. The branches and dissepiments are slender, and more nearly of the same size than is usual, the branches being naturally somewhat larger. The fenestrules are slightly longer than wide, being angular on the nonporiferous side and more rounded on the other side. The dissepiments are slightly depressed.

The outer surface of the branches is marked by fine, obscure, interrupted striæ. On the poriferous side they are traversed by a high carina, not very greatly expanded at the top, which is wavy and nodose. Along the center of each is a row of little spinules, which are often arranged so that one is opposite each dissepiment and two are intermediate. Opposite each dissepiment is an aperture, while two others are intermediate along the sides of the fenestrules.

F. texana has a fine regular mesh, like several of the other forms described, but is distinguished at once by the slenderness and nearly equal size of the dissepiments and branches and the nearly square shape of the fenestrules. Other differences also appear in individual cases.

This species is probably new, but I have hesitated, because of the small amount of my material and the imperfect information regarding some of its details, to give it a new name. It is related to Fenestella binodata, F. corticata, F. delicatula, F. modesta, and F. subretiformis, but is more or less strongly distinct from any of them. It is perhaps especially close to F. modesta, but differs in its more regular growth, in having a more elevated carina, and possibly in other characters.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

FENESTELLA sp. a.

Pl. XXVIII, fig. 6.

A small specimen, about 5 mm. square, is all that our collection contains of this species. It is silicified and at present free from adhering rock, but some of the most important characters have been lost. So far as may be judged from material in hand the zoarium was a large, flat, regular frond.

The growth is very uniform. Five rows of fenestrules and their inclosing branches (or a little less) occur in 3 mm. transversely, and four fenestrules with their inclosing dissepiments occur longitudinally in the same distance. The fenestrules are nearly twice as wide as long. They are elliptical in outline on the poriferous face and somewhat quadrate on the opposite one. The dissepiments are one-half the width of the branches, or in some cases less, and usually lie even with their surface on the nonporiferous side, but are depressed on the poriferous side. The back of the frond appears to be entirely smooth, but delicate sculpture may have been lost in the process of silicification. This process has apparently obscured the zoœcial structures, whose number and character can not now be determined. There appears to have been a carina, but details as to its height, etc., can not be ascertained.

This form resembles several found in the Pennsylvanian strata of the Mississippi Valley, but it is doubtful if it prove to be the same specifically, even if all its characters were known, for those which have been determined do not entirely agree with any of them. It is as near Fenestella remota as any, but that species is striated on the reverse side. F. parvipora is another related species.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

FENESTELLA sp. b.

Pl. XXVIII, fig. 10.

This species is similar to that discriminated as Fenestella sp. a, but has a distinctly coarser mesh.

What may be taken as the typical example shows four fenestrules transversely and three longitudinally in a space of 3 mm., with the inclusive branches and dissepiments. The fenestrules are about twice as long as wide, and the branches about twice the diameter of the dissepiments. On the nonporiferous side the dissepiments are on the same plane as the branches, but they are depressed on the poriferous side. The fenestrules are strongly quadrate on the nonporiferous side, but more elliptical on the other.

The nonporiferous side appears to be without ornamentation of nodes or striæ. The character of the poriferous side has been obscured during silicification, so that even the number of apertures can not be made out.

This form resembles in a general way several species found in the Pennsylvanian of the Mississippi Valley, especially F. binodata and F. subrudis; but the final comparisons for determination of the actual degree of difference and resemblance can not be made. It may be no more than a coarse variety of that discriminated as Fenestella sp. a.

In addition to the fragment on which the foregoing description is chiefly based, several others having essentially the same character have been placed in this species.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

FENESTELLA sp. c.

Pl. XXVIII, fig. 7.

This form in many respects resembles the foregoing (Fenestella sp. b), but differs in several points. While constructed on about the same scale and with about four rows of fenestrules in 3 mm., the fenestrules themselves are a little more elongate, there being only from two and one-half to three in a space of 3 mm. The growth is considerably less regular. The dissepiments are often depressed on the nonporiferous side. The fenestrules are oval. This side of the frond (the nonporiferous) appears to be ornamented with fine longitudinal striæ not shown in my figure, but this may be due to imperfect silicification of the fibrous skeleton.

The poriferous side is not satisfactorily preserved, being more or less eroded, but the apertures seem to have numbered about three to a fenestrule.

This species is related to several from the "Coal Measures" of the Mississippi Valley, especially to F. binodata, F. kansasensis, F. modesta, F. ovatipora, and F. subrudis, but especially resembles F. modesta. The preservation is, however, scarcely competent to a careful determination of its specific relations.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains Texas (station 2969).

FENESTELLA sp. c. var.

This form is represented by a single specimen whose characters bring it close to Fenestella sp. c., but as it was obtained from a different locality and presents some differences of importance it did not seem advisable to refer it to the same species without further evidence of intergradation.

The poriferous side of this specimen is uppermost, the opposite side being embedded in rock. Exfoliation has removed all superficial characters of the exposed portion. The branches are straight and about equal distances apart. The dissepiments, however, stand at varying intervals front one another, so that the fenestrules range in length from two and one-half to one and one-half times their own width. They are subquadrate in shape. The dissepiments are slender and depressed, in width about half that of the branches, which are in turn about half as wide as the fenestrules. Five rows and six branches come in 3 mm. transversely and two and one-half fenestrules in 3 mm. longitudinally. Four cells, or in some cases three, occur opposite each fenestrule.

The chief differences manifested between this form and typical Fenestella sp. c are the subquadrate shape of the fenestrules (which may be due to their being viewed chiefly on the opposite side), their slightly narrower width, so that more rows occur in a given distance, and the larger number of cells which lie opposite each fenestrule. It appears to be closely allied to Fenestella sp. b also, perhaps more nearly than to the present form.

Horizon and locality.—"Dark limestone," hill southwest of Guadalupe Point, Guadalupe Mountains, Texas (station 2924).

FENESTELLA sp. e.

Pl. XXVIII, figs. 5 and 5a.

Like the foregoing, this is a small species and discriminated on very fragmentary material. The branches are straight and extremely heavy. The dissepiments, on the other hand, are as a rule short and thin. The fenestrules are elongate elliptical. Sometimes the dissepiments are much thicker at the expense of the fenestrules which they separate, but in any event, being sunk beneath the level of the thick, strong branches, they play in appearance a subordinate part in the construction of the zoarium. There are about three rows of fenestrules and four branches, or a little over, in a distance of 3 mm. Longitudinally, there are three fenestrules (and four dissepiments), or a little less, in the same distance.

The nonporiferous side is without much doubt artificially varicose from heavy silicification, the original ornamentation, if such there was, being concealed. On the obverse side the apertures are arranged about as in Fenestella hilli, namely one opposite each dissepiment and two intermediate. There is in addition a high carina, much expanded at its top, upon which are set, in regular arrangement, a number of large nodes or spinules. These occur about six to a fenestrule, arranged alternately, three on each side of the broad summit of the carina, to which their enlarged bases, practically in contact, lend a sinuous outline. It may be that these structures also have been exaggerated by heavy silicification, but whether this is so, or to what degree, it is impossible to say. I think, however, that the real appearance may have been appreciably altered by exaggeration in this manner.

This species also is probably new, but for the reasons mentioned in the foregoing case a distinctive name has not been applied to it.

It is related to several forms which have been described from the "Coal Measures" of the Mississippi Valley and in the proportional measurements to Fenestella dentata and F. kansasensis; but from these it differs considerably in other details. In the structure of the carina a parallel form is found in F. binodata, but there are minor differences in the carina itself besides those of proportion, which forbid joining the Guadalupian form with that species. In the Guadalupian itself we have F. popeana, which shows much the same dimensions, but possesses an obsolete carina instead of a highly developed one. Fenestella subretiformis described from the Organ Mountains is related in a somewhat similar manner, and besides possessing important structural differences is believed to hold a much lower horizon than the form under consideration.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

FENESTELLA sp. f.

Pl. XXVIII, fig. 8.

This species resembles the foregoing in having the branches relatively massive for the dissepiments, but is somewhat more delicate in construction. There were probably four rows of fenestrules and five branches in 3 mm. and about five fenestrules and five dissepiments longitudinally. The fenestrules are elliptical and from one and a half to two times as long as they are wide. The dissepiments are somewhat depressed on both sides of the frond and the fenestrules elliptically elongated. I am not sure whether the nonporiferous side was ornamented or not, but the branches appear to have been somewhat nodose at their junction with the dissepiments, and there are strong indications of other nodes of smaller size. On the poriferous side there was probably a high carina, whose character can not be made out in detail. The zoœcial apertures are so arranged that one of them is placed opposite each dissepiment and one in an intermediate position.

The disparity between the branches and the dissepiments is less marked in this species than in the foregoing, and the proportions, as well as the arrangement of the zoœcia, are also different.

Of this form too little is known to determine its relationship to existing species, but it appears to be as yet undescribed. It shows some resemblance to Fenestella shumardi and F. spinulosa, but is clearly distinct from them. One of its most peculiar features is the large number of fenestrules relative to the number of rows in a given distance. This relation is not very well shown by the specimen, however, and the statement made above may represent an extreme case.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

FENESTELLA sp. f?

The specimen referred to by this title very closely follows the characters of the typical variety save in one important particular. There are four rows and four branches in 3 mm. transversely and four fenestrules and five dissepiments in the same distance longitudinally, these particulars being thus very similar to but not identical with those of the type. The dissepiments are much depressed on the nonporiferous side and very thin—one-third to one-fourth the diameter of the branches. The latter are not as strong as in the type, but are as wide or slightly narrower than the fenestrules. The latter average one and one-half times as wide as long and have a more or less quadrate shape.

The cells are so arranged that one is opposite each dissepiment and one opposite the center of each fenestrule. They are separated by a high carina, expanded at its top.

The nonporiferous side is marked by fine, linear, longitudinal striæ and possibly by nodes, though the presence of the latter is not a positive character.

The most marked difference which seems to separate this form from Fenestella sp. f is the striated surface and the much less numerous and less obviously developed nodes. There are perhaps very faint traces of striæ in the latter form which may have been obscured by coarse silicification, and the fact that it probably represents an older portion of the zoarium may account for the more strongly developed nodes.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Genus POLYPORA McCoy.

This genus, somewhat in contrast to Fenestella, is better represented, in both species and individuals, in the Salt Range than in the Guadalupian fauna. Waagen and Pichl recognize eight species, some of them fine, robust types and represented, it would appear, by large, well-preserved fronds. My material shows representatives of what may be four species; but these occur for the most part as small fragments, often silicified or otherwise poorly preserved. Where it seemed to be demanded, comparison has been made with Pennsylvanian forms from the "Coal Measures" of the Mississippi Valley; but the character of my fossils seems scarcely to warrant careful comparisons with Salt Range species, as such comparisons could hardly be anything but unprofitable.

POLYPORA MEXICANA Prout?

Pl. XIX, figs. 6 to 6b.

1858. Polypora Mexicana. Prout, Trans. Acad. Sci. St. Louis, vol. 1, p. 270 (date of volume, 1860).

Permian: Jornada del Muerto, New Mexico.

1859. Polypora Mexicana. Prout, idem, p. 451, pl. 16, figs. 2-2b.

Of this species but three specimens have come to hand, two of them mere fragments. While it has been possible to free them from the inclosing matrix so that both surfaces are presented to view, the silicification which rendered this practicable is not so perfect but that some of the more delicate details have been obscured or misrepresented.

The zoarium in this specimen is rather large and heavily constructed, and shows irregularity in many details. Instead of being planate or infundibuliform, it is strongly undulated or contorted. The branches are of nearly equal size, except at points of bifurcation; but bifurcation seems not to occur at regular or rhythmic intervals. The dissepiments, on the other hand, when viewed on the poriferous face, vary in size from equality with the branches to a width one-half or one-third as great. The fenestrules, while as a rule of rather constant width, vary much in length, and range in shape from subcircular to elongate. The zoœcial apertures are rather evenly distributed, but are not conspicuously arranged in rows, either vertical or oblique, but of the two a vertical arrangement is more obvious. In some cases the dissepiments are celluliferous and in others not. The branches which are rather thick through, are slightly narrower on the nonporiferous than on the poriferous face; but this is far more marked in the case of the dissepiments, which appear much thinner and are depressed below the level formed by the back of the branches.

Measured on the poriferous face there are from two to two and one-half fenestrules (and three dissepiments) in a space of 5 mm.; measured transversely there are from two fenestrules and three branches to four fenestrules and four branches in the same distance.

The zoœcial apertures are small and separated by intervals more or less nearly equal to their own diameter. The number lying opposite a fenestrule naturally varies considerably with the size of the fenestrule, but it seems to range from three or four to six, without counting those immediately opposite the dissepiments. They come in four to six or seven longitudinal rows, but, as before remarked, this arrangement is not very conspicuous. Both faces of the zoarium appear to be without ornamentation, except that the lower portion of the rim of many of the apertures seems to project as a little lip or spine; but this may be merely an effect of preservation. The nonporiferous side appears to be quite smooth.

A large number of species have been considered in the effort to identify this form, but none of them possesses more lines of approximation than Polypora mexicana Prout, which, moreover, was described from the same general region, though from a horizon which at present appears to be somewhat lower. The points of resemblance are sufficient to have encouraged me to make the identification; the points of difference have caused me to feel some doubt of it.

The chief differences are as follows: The zoarium in the typical specimen is regular in its curvature and structure, while the form under consideration is contorted, its fenestrules are of unequal size, and its structure in general is somewhat more irregular than that represented in Prout's figures and description. While radial measurements in the two specimens give the same result, the present example seems to be a little more expanded in a transverse direction. The size of the zoœcial apertures and their number in a given interval appears to be about the same; but in the type the apertures are said to be regularly distributed in oblique rows, an arrangement not often very obvious in my form. Prout, furthermore, represents the apertures as emerging from little elevations, the bases of which are well nigh in contact, the actual apertures, on the other hand, being farther apart than in my specimen. It is doubtful if the details of this face were sufficiently preserved in the original example to bear out Prout's figures, especially as he does not describe this appearance in the text. Several important features are left unknown in Prout's description, especially those of the nonporiferous face; and it is quite possible that the differences already noted are indices of still greater ones that would appear were the full details of both forms available for comparison.

Prout cites this species from the Permian and gives as a locality the Jornada del Muerto. It is very doubtful if the Guadalupian occurs in that region, the real horizon of P. mexicana being, therefore, the Hueco formation. This constitutes an additional reason for doubting the identity of the present form with Prout's species.

Several American species resemble Polypora mexicana more or less closely, but besides showing differences of character they are from distant areas and from different faunal associations. One of these is P. burlingtonensis, a species in fact closely similar in a general way but more regular in structure and less sturdy in growth. P. simulatrix also shows some resemblances, but is not sufficiently close to make comparisons necessary. Of much nearer relation, however, is the imperfectly known P. crassa, which differs at least in this particular, that the fenestrules are regularly more elongate. Of the species from our western States, aside from that with which the specimen under consideration has been identified, none possesses characters sufficiently close to admit confusion. The same is true of most of the Indian species described by Waagen and Pichl, many of which are characterized by very robust growth. One of the most similar is doubtless P. sykesi De Koninck, which is readily distinguished, however, by its large size, more regular growth, less elongate fenestrules, and heavier dissepiments. The same differences distinguish P. ornata, which, moreover, possesses a system of ornamentation unknown in the American species.

De Koninck a has placed P. mexicana in the genus Protoretipora, I know not on what evidence, but he is clearly mistaken if the present identification is anywhere near correct.

---

aMem. Geol. Survey New South Wales, Pal., No. 6, 1898, p. 137.

---

Horizon and locality.—"Dark limestone," Pine Spring (station 2930) and Guadalupe Point (station 3762e), Guadalupe Mountains, Texas.

POLYPORA sp. a.

Pl. XXVIII, figs. 1 and 1a.

Of this form our collection contains a fragment, which is well preserved on neither side. It is a moderately robust form, of fairly regular growth. The branches are straight and strong, highly rounded on the reverse side. They expand considerably at their junction with the dissepiments, which are narrow in the middle and also considerably expanded. The dissepiments are much depressed.

The fenestrules are elliptical in outline, about twice as long as wide, and with a width slightly greater than that of the branches. There are four rows (and five branches) in a space of 5 mm., and three to three and one-half fenestrules (and four dissepiments) longitudinally in the same distance. The surface of the reverse is rendered rough by the nodular silicification, but there appear to be traces of an originally striated sculpture. The present nodular surface is probably not an original character.

The surface of the obverse base is eroded, so that the cell walls are exposed. Consequently its superficies can not be determined. The cells in this preservation are rhomboidal and regularly arranged. At present the diagonal instead of the longitudinal rows, of which there are usually four, are most noticeable. There are about four opposite each fenestrule. Sometimes one stands opposite a dissepiment, with three intermediate.

This species is of finer and more regular growth than the form referred to, Polypora mexicana?, a specimen of which is found at the same locality, and it probably has a striated instead of a smooth reverse side. The latter character, if it prove a real feature instead of an appearance due to silicification, would aid in discriminating this form from several Mississippi Valley species which it resembles. In this category may be mentioned P. bassleri, P. cestriensis, P. distincta, P. stragula, and P. ulrichi. These are all more or less closely related, but even if the sculpture of the reverse of the form under consideration is found to be illusory it hardly agrees exactly with any one of them, and is likely to prove a new species.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

POLYPORA sp. b.

Pl. XXVIII, fig. 2.

A number of fragments belonging to this species have come to hand, but they are all small and poorly preserved. They show an irregular and coarsely reticulate form, the branches of which are moderately stout, flexuous, and frequently bifurcating. When, at unequal intervals, the branches come close to one another, they are connected by short but moderately strong dissepiments. Often the resulting network looks as if through their irregularities the branches have actually become connate without any connecting dissepiments.

The branches are strongly rounded on both sides, so that a cross section is in a general way circular. The superficial character of both surfaces has been altered by imperfect silicification. The reverse is irregular and nodose, due, I feel no doubt, to chalcedonic replacement, and a similar feature may be noted in some cases on the obverse. The latter surface is thickly covered with apertures, which have no well-defined interspaces and are disposed in five rows, sometimes a greater number. Their linear arrangement is not striking, and that in an oblique direction is possibly more noticeable than in a longitudinal. They appear to open obliquely and to have an elliptical outline, the interspaces, which seem to represent the thickened zoœcial walls, being about the width of their shortest diameter. The general appearance of these zoœcia resembles irregular growths of Acanthocladia guadalupensis, whose branches are connected by dissepiments.

Owing to the large size of the mesh and the small size of the fragments it is impossible to give measurements in the ordinary way. Large branches have a diameter of about 1 mm. or a little over, but there have been referred here fragments, supposed to be terminal in position, which are considerably more slender. A long fenestrule is about 3 mm. in length, the width, which is usually about that of the intervening branches, being about one-third. In shape the fenestrules often taper almost to a point at one or both ends.

This species appears to be nearest related to P. crassa, of the "Coal Measures" forms of the Mississippi Valley, but both species are too imperfectly preserved to render possible a determination of their relationship. It is unlikely that they are specifically the same.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

POLYPORA sp. c.

Pl. XXVIII, fig. 3.

This species is represented by a very small fragment, whose characters ally it in some degree to Polypora sp. a. It is, however, a considerably coarser form. There are about three rows and three branches in 5 mm. and about two fenestrules and one dissepiment in the same distance. The branches are about 0.75 mm. in width, straight, somewhat expanding at their juncture with the dissepiments, slightly narrower than the fenestrules. They are flattened and apparently striated on the reverse and rounded on the obverse. The dissepiments are slender (about one-third the branches), somewhat depressed, expanding. The fenestrules are subelliptical to subquadrate, about twice as long as wide.

The apertures lie in about seven rows, the arrangement being more noticeably diagonal than longitudinal. They are about their own diameter apart. Five are opposite a fenestrule, with sometimes an additional one opposite a dissepiment.

In addition to the small silicified fragment on which the foregoing characterization is based, there has been provisionally referred here an equally fragmentary example from station 2930, embedded in limestone, with the poriferous face exposed. The network is of about the same degree of coarseness as the fragment whose characters have just been set down, but the celluliferous face, which probably comes from one of the early or initial portions of the frond, has been so thickened with testaceous deposit as to present characters probably quite different from those which really belong to the species. It is possible that these two specimens have been incorrectly grouped together.

This form answers very well to Prout's description of P. mexicana. It differs, however, from the form here provisionally identified as that species in its more regular growth, thinner dissepiments, and somewhat coarser reticulation. It may prove to be the same, but this is rather unlikely.

It would appear to be not unlike P. remota, though showing marked differences in some respects, such as the much closer arrangement of pores on the obverse and the striated sculpture on the reverse. It must be conceded, however, that in both these particulars the imperfect preservation of the Guadalupian species may have led to misinterpretation of the real characters. Another similar species is P. submarginata, but it likewise shows distinct differences in certain points, such as the central row of nodes on the reverse.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930?). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

POLYPORA sp. d.

Pl. XXVIII, fig. 4.

Two small and poorly preserved fragments are all that represent this species, of which it is impossible to give a description at this time. In most of its characters it rather closely resembles the form here provisionally placed under Polypora mexicana, but the reticulation is distinctly coarser and the fenestrules of very unequal sizes. In the proportions of its reticulation and in some particulars it resembles the form designated as Polypora sp. c, but besides being somewhat more heavily constructed, a difference which under the circumstances would not necessarily be very important, it is much more irregular in the size and shape of its fenestrules, which are also, as it were in compensation for the heavier branches, somewhat smaller. It may be that the species c and d will prove to be the same, since the fragments of both forms at present known are so small as to hardly justify an unqualified inference that they are representative of the entire zoarium.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Genus PHYLLOPORA King.

PHYLLOPORA ? sp.

Our collections contain but one fragmentary specimen of this form, regarding whose generic position I am in some doubt whether it belongs to Phyllopora or to Polypora. The branches, which seem to be fairly straight, have a width of 1 mm. or a little less, and this is slightly greater than the width of the fenestrules. The latter are two to three times as long as wide, elliptical, narrowing strongly at either end. What would necessarily be regarded as dissepiments if the form is a Polypora are well nigh as broad as the branches, sometimes depressed, sometimes celliferous. The zoœcia appear to occur in four to six rows. The nonporiferous side is without ornamentation.

From the small silicified fragment which we have of this species it is impossible to place it satisfactorily. The branches are almost too distinct from their connections for Phyllopora, while the dissepiments are almost too broad and indistinct as such to make this a typical Polypora.

Horizon and locality.—Delaware Mountain formation, mountains northwest of Marathon, Tex. (station 3840).

Genus THAMNISCUS King.

THAMNISCUS DIGITATUS n. sp.

Pl. XXV, figs. 6 and 6a.

This species forms small, branched, more or less palmate colonies, which are attached by a short peduncle. Several branches spring from the peduncle, and these themselves divide, especially a short median main branch, which gives off a few secondary branches pinnately, the whole growth being in a general way irregular yet not especially unsymmetrical. The branches are about 2 mm. in diameter, stout, short, and dactyloid. All bend upward, forming a shallow cup-shaped colony about 13 mm. long and 11 mm. wide. Along the center of the upper side of each branch extends a band of small zoœcial openings which appear to be in about six longitudinal rows and to have more or less of an oblique arrangement also. As well as can be counted seven occur in a distance of 3 mm., longitudinally. In mature or old colonies the zoœcial band is strongly elevated and narrower than the branch; occupying only part of its upper surface. Toward their bases the branches are still more expanded, spreading out on each side into a sheet, which becomes confluent with similar expansions of adjoining branches. The back of the frond appears to be smooth.

Several specimens of this species have come to hand among the silicified material from the Diablo Mountains. Although I originally supposed that they might be old specimens or basal portions of colonies of Acanthocladia guadalupensis, this now hardly seems to me a tenable hypothesis, and I have decided to introduce a new name for them, referring them to the different but kindred genus Thamniscus.

This form seems to be closely related to Thamniscus palmatus of Condra. A number of differences can be named between my typical specimen and Condra's description and figures, but the other Guadalupian specimens fail to show some of the characters of the type, which is probably an old colony. It differs from T. palmatus somewhat in its manner of developing the zoœcium, but principally in having much thicker branches and with expanded and confluent basal portions.

Horizon and locality.—Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764).

THAMNISCUS sp.

Pl. XXXI, fig. 16.

Of this species our collection has furnished but a single example, which is distinguished by its delicate construction. The branches have a diameter of only 0.5 mm., and they divide at intervals of about 3 mm. The nonporiferous side appears to have been smooth and the zoœcia on the poriferous side seem to have been arranged in four to six longitudinal series, but the silicification which these specimens have undergone has more or less obscured the character of both surfaces.

This is without much doubt a species new to the American Carboniferous at least, but it seems that its characters should be determined on better and more extensive material before it is described under a new name.

Horizon and locality.—Delaware Mountain formation, mountains northwest of Marathon, Tex. (station 3840).

Family ACANTHOCLADIIDÆ Zittel.

Genus ACANTHOCLADIA King.

ACANTHOCLADIA GUADALUPENSIS n. sp.

Pl. VIII, fig. 1; Pl. XVIII, figs. 13, 13a, 14, 14a, 16 to 16b; Pl. XXII, figs. 10 and 10a.

?1859. Acanthocladia Americana. Shumard (non Swallow), Trans. Acad. Sci. St. Louis, vol. 1, p. 388 (date of volume, 1860).

Gray [Permian] limestone: Guadalupe Mountains, Texas and New Mexico.

This is one of the most abundant Bryozoa of the entire Guadalupian fauna, and numerous examples have been examined. The following description is derived from the typical specimen, which was obtained at the horizon of Shumard's "dark limestone." Much of the material is less complete and less perfectly preserved than the typical specimen, and shows certain departures from it to which reference will later be made, but which do not enter into the specific diagnosis.

Zoarium robust, consisting of three somewhat sinuous branches lying in the same plane. The two lateral branches are nearly opposite and directed at an angle of about 60° or less to the median one, which is persistent. The branchlets make, with the branches, a nearly flat frond, but stand at various angles to them. Usually the angle is slightly less than 90°, and it varies to about 60° The branchlets also present considerable variety in length, the longer ones being themselves branched. They are a good deal smaller in diameter than the branches; otherwise the forked branchlets might themselves be considered branches. The branchlets attain an extreme length of 5 mm., but usually are shorter. The shortest, which have without much doubt been broken, are just prominent enough to be recognized. Probably 3-1/2 or 4 mm. would cover the length of the majority of perfect ones. The average width of these is from 0.75 to 1 mm. The thickness of the branches ranges from 1-1/2 to 2 mm.

On the poriferous face of the zoarium, especially on the branches, the apertures are grouped along a central zone, which is usually much elevated. The expansion of the branches below the celluliferous portion seems to be a rather striking character in all the specimens seen. On the branchlets this does not occur or is much less obvious. On this portion of the frond the apertures extend far down on the sides. They are small, only one-sixth to one-eighth of a millimeter across, and open somewhat obliquely lengthwise of the axis, so that their shape is rather elongated. The distal margin projects upward into a spiniform lip, which in well-preserved specimens like the type is a striking feature. The spines cover the poriferous surface, which they ornament in a beautiful manner. Those on the sides appear to be longer than the rest and tend to form a lateral row of especial prominence. The apertures are separated from one another by rather thick walls, their average distance apart being about the same as their average diameter. On the poriferous face I have been able to detect no auxiliary pores, no raised or depressed lines, pustules, or ornamentation other than the spines, which abundantly cover the surface. The apertures are not obviously arranged in rows, but appear for the most part rather promiscuously crowded together. The eye follows them along diagonal rather than longitudinal lines, of which, loosely considered, there may be four or five, with a slightly narrower band on the branchlets.

On the reverse side the branches are well rounded. The branchlets, also fairly round, are much depressed. The nonporiferous surface of the typical specimen is obscured by the coarse plates of chalcedonic silicification, but another example, so closely associated with the typical one that I am not sure that they did not belong to the same zoœcium, shows the reverse side to be covered by very delicate, moderately strong, somewhat flexuous and inosculating longitudinal liræ, which gives this surface an exceedingly handsome appearance. These markings are too fine to be represented in my figures.

I have identified this species at a number of localities, and referred to it most of the material belonging to the genus which has come into my hands. In most cases the fossil is silicified, and by etching is obtained in a free condition. While this is in some ways an advantage, the silicification is often rather coarse or chalcedonic and the specimens are fragmentary. Considerable variation is shown. A rather constant character is the presence of a broad elevation along the median portion of the main branches, to which the zoœcial apertures are restricted. While the distance between the pinnules and their general arrangement remain fairly constant, considerable difference in size and some in appearance is produced by age. As the zoarium grew older it increased in size, not only in a lengthening of the branches (the pinnules apparently remaining more nearly constant) but also in a thickening of them, chiefly over the back. Layer after layer was added to the nonporiferous side, making a thick, strong basal plate, and at the same time producing a gibbous arching of the surface which elevates it far above the plane of the pinnules. In specimens which are not silicified these plates of shell peel off in concentric exfoliation. Thickened specimens sometimes reach a diameter much greater than that of the types and indicate that these colonies attained a considerable size. Another effect apparently produced by thickening was to straighten the branches, which near their terminations are slender and have a zigzag course in relation to the development of the pinnules. Whether it also caused the pinnules to appear less distinctly alternating than is perhaps the rule, I am in doubt. Several examples occur, however, in which these branchlets originate nearly opposite each other. To the same cause may perhaps be attributed the absence in some specimens of the fine striation which was seen on the back of one of the types. Other specimens from the same station fail to show this sculpture, and it is conceivable that it is not developed on later testaceous deposits. A character which occasionally appears in thickened examples consists of a few irregularly distributed nodes along the median line of the back of some of the pinnules. (See fig. 16a, Pl. XVIII.)

The poriferous face also appears to have received testaceous deposits, though to a less extent than the other. The tendency in this case appears to be the closing of the zoœcial openings and the obscuring or burying of the spines which project from them. While as a rule the branches and branchlets lie in nearly the same plane, in some examples the pinnules bend backward so that the nonporiferous faces make an angle of less than 180°. Under such circumstances the pinnules do not of course appear depressed on the reverse side.

The little spines, which are a striking feature of the poriferous surface of the type specimen, have been observed only here and there, and the fine liræ seen on a specimen accompanying the type in no other instance. Much of the material seems to have been more or less worn, the pinnules occasionally reduced to short stumps cut obliquely across. This usage would of course tend to destroy primarily the little spines on the front and also the striæ on the back of the zoœcium. Silicification has also done its part in affecting the appearance of these organisms. Usually it is of a chalcedonic nature and tends to give them a wrinkled or warty look and to obscure all structures. Any delicate sculpture on the back would almost certainly be obliterated. On the front it operates to close the zoœcial openings and generally to obscure them, perhaps sometimes to replace them by small spinelike projections, though I think that the spines on the typical specimen are not due to this cause. It is possible, however, that the very delicate striation noticed on the auxiliary specimen, but in no other instance, may have resulted from this process. While as a rule more or less obscured, some silicified specimens were almost certainly smooth on the nonporiferous side.

Thin sections, where the preservation is good, show that the test is very finely tubular. The tubules are generally normal to the outer surface, but their direction from the zoœcial apertures is radial.

I am not altogether satisfied that all the specimens identified with Acanthocladia guadalupensis which show the variation mentioned above belong to the same species as the type. Could I be sure that the differences noted are not due to imperfect preservation and to unequal age a separation into several species would seem to be demanded, though under present conditions impossible to be satisfactorily carried out. On the other hand, the factors tending to produce such apparent differences are certainly operating, and it has seemed to me more appropriate to refer all to one species.

Acanthocladia guadalupensis is so unlike A. americana, A. fruticosa, A. pinnata, A. anceps, or any of the species known to me that a detailed comparison is unnecessary. I may recall, however, that Shumard suggested the name americana for a form from Kansas, and that a year later he identified the same species in the Guadalupe Mountains. It seems altogether likely that the Guadalupian form which he refers to A. americana is that described here. There is little likelihood of A. guadalupensis occurring in the Mississippi Valley, and little also of its being the same as A. americana, which though imperfectly described is nearly related to A. anceps Schlotheim.

Horizon and locality.—Top of Capitan formation, Capitan Peak (stations 2966? and 3762a); middle of Capitan formation, Capitan Peak (station 2926): base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930), hill southwest of Guadalupe Point (station 2924), Guadalupe Point (stations 3762b, 3762c, 3762d, and 3762e); Delaware Mountain formation, Guadalupe Point (station 2903), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2957, 2962, 2969, and 3500). Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764). Delaware Mountain formation, Comanche Canyon, Glass Mountains (station 3763), and mountains northwest of Marathon (station 3840), Texas.

ACANTHOCLADIA sp.

Pl. VIII, figs. 2 and 2a; Pl. XVIII, figs. 15 to 15b.

I have been unable to refer to Acanthocladia guadalupensis several fragments found in the Capitan formation and the "dark limestone." The size and mode of growth of this form are not well known, but apparently do not differ materially from those of the common species. The cells do not, however, open from a central raised zone. In fact, the poriferous surface is rather markedly flattened, and the apertures distributed all over the obverse side of the branches and branchlets. The intervals between them, furthermore, are wider than in the common type, and the arrangement into rows, especially diagonally, is a little more obvious. The general appearance is not unlike that of Acanthocladia anceps Schlotheim. It is possible, however, that both these peculiarities—the absence of a raised zone and the wider spacing of the apertures—may be the result of a deep exfoliation of the poriferous face. I hardly think, however, that this has occurred. The nonporiferous side, in the single instance where it has been possible to observe it, is quite smooth; but, as in the case of other silicified specimens, possible ornamentation of some sort has been lost, either by erosion or during the process of silicification.

Horizon and locality.—Capitan formation, Capitan Peak (station 2926); "dark limestone," Guadalupe Point (station 3762e), Guadalupe Mountains, Texas.

Genus SEPTOPORA Prout.

SEPTOPORA aff. S. ROBUSTA Ulrich.

Of this species but two specimens have come to hand and the fragments are so small and the silicification so imperfect that the proposal of a new name hardly seems justified.

The form under consideration is characterized by the slender branches, which have about the same diameter as the dissepiments, and by the quadrate proportions of the fenestrules. Two branches and one fenestrule seem to come in 2 mm. in a transverse direction. The branches have a diameter of half a millimeter and the dissepiments are as a rule equal to them. The fenestrules vary from somewhat longer than wide to somewhat wider than long, and the shape is quadrate.

It is difficult to give data regarding the zoœcia from my material. They seem to occur in two rows, with a slight carina between, but sometimes there are three, especially near the junction with the dissepiments. Three or possibly four zoœcia occur opposite each fenestrule, with two opposite each dissepiment.

The nonporiferous side seems, to be without ornamentation, but is pierced by rather large, somewhat widely spaced auxiliary pores, more or less irregularly ranged in three alternating series.

This form appears to be quite closely allied to Septopora robusta of the Pennsylvanian, but my material is too imperfect to permit a comparison in all points.

Horizon and locality.—Delaware Mountain formation, mountains northwest of Marathon, Texas (station 3840).

Family RHABDOMESIDÆ Vine.

Genus RHOMBOPORA Meek.

RHOMBOPORA aff. R. LEPIDODENDROIDES Meek.

Pl. XXXI, fig. 17.

To determine satisfactorily just what are the characters and affinities of Rhombopora lepidodendroides Meek would be a very difficult matter, seeing that the typical specimens can not be definitely fixed upon and the ramifications of the subject are extensive and intricate; but it can at least be said that the form under consideration is extremely similar to Meek's figures in his report on the paleontology of eastern Nebraska.

The fragmentary example which alone represents this species, having a length of 5 mm., is but 1-1/2 mm. in diameter. The cylinder thus defined is crossed externally by two sets of strongly oblique ridges going in opposite directions, marked with nodes where they meet and with granules on the intervening portions. The rhombic depressions which they form, constituting the vestibular portion of the zoœcial tubes, narrow down in such a way that the apertures of the zoœcia are elliptical, very narrow, about two or three times as long as wide. About three rhombs or a little less occur longitudinally in 1 mm.

It has been impossible to study this form by means of thin sections, since only a single specimen has been found, and that is silicified. In view of the fact that neither form can be said to be satisfactorily known, I feel indisposed to place the Guadalupian one unreservedly with Rhombopora lepidodendroides, especially since the associated faunas are so unlike, and since certain differences appear to exist between the Guadalupian specimen and Meek's figure, such as the greater obliquity of the rows of cells and the more elongate shape of the zoœcial apertures.

Horizon and locality.—Delaware Mountain formation, mountains northwest of Marathon, Texas (station 3840).

RHOMBOPORA? sp.

Under this title are included three specimens from two localities, which appear to belong to the genus Rhombopora. They are small fragments and are silicified. The range of diameter of the branches is from 2 to 3 mm. The smallest specimen has the cells obviously arranged in diagonal rows, in which they are about their own diameter apart. Longitudinally they are a little more than their own diameter apart, and about four cells and four intervals occur in a distance of 2 mm. The zoœcia are circular, and the vestibule appears to have essentially the sane shape; at least it is not conspicuously rhombic or elliptical, nor does the outer surface appear to have been studded with tubercles. If Rhomboporas, therefore, these hardly belong to the lepidodendroides group. In the larger specimens the spiral arrangement of the zoœcia is much less rigid and extensive, which, taken with some slight deviation in their size, suggests that the larger examples belong possibly to another species, or even to another genus.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2957 and 2969).

Family CYSTODICTYONIDÆ Ulrich.

Genus GONIOCLADIA Etheridge.

GONIOCLADIA AMERICANA n. sp.

Pl. VIII, figs. 3 to 3c.

Waagen and Pichl relate that they had but a single specimen with which to carry on their studies of Goniocladia indica, and a supply of material equally limited circumscribes my observations on the American form. While I have little doubt that the latter is distinct from the Indian species, as well as from G. cellulifera, the only other member of the genus known to me, it is with some hesitation that I have described it as new, because of the rather imperfect knowledge which it has been possible to obtain of it.

Goniocladia americana consists of a more or less flat expansion formed by interlocking branches, producing a more or less regular network. In the single specimen known it is impossible to trace the constituent arms into continuous branches. Still less is it possible to distinguish any arrangement of branches and dissepiments. The thickness of the branches is as a rule considerably greater than their width from side to side. The nonporiferous face is sharply angulated, the surfaces which slope away from the central line being more or less flattened. Aside from this angulation, the nonporiferous side seems to be without ornamentation. The poriferous side is not shown by the specimen, but appears to be strongly rounded without any distinct carination. There seems to be a median plate from which the cells diverge. It is quite distinct along the nonporiferous side of the frond, but has not been seen on the poriferous half of the structure. Near their point of origin the cell walls are rather thin, but they rapidly become thicker with advancing length, so that their apertures on the poriferous side are separated from one another by intervals considerably greater than their own diameter. The width of the branches is frequently 2 mm. and sometimes 2-1/2 mm.; it is sometimes also less than 2 mm. The fenestrules vary so greatly in size and shape that it is difficult to make a statement in regard to them which will be at once specific and true. They seldom have a length of 6 mm. and are frequently smaller. The width of the fenestrules is still more variable than their length, though they are seldom as long as wide, and sometimes are very narrow.

The characters which seem satisfactorily to distinguish the American form from that described from India are these: Our species is distinctly more robust, with heavier branches and larger fenestrules. The outlines of the fenestrules are not so strikingly serrated as shown by Waagen and Pichl's figures, though possibly no great difference exists in this particular. The poriferous side of the American form appears not to be carinated, though possibly on this point also but little stress should be laid, as my observations are imperfect and unsatisfactory. G. americana also apparently lacks the striated surface ornamentation of the Indian form. Waagen and Pichl do not describe the great thickening of the cell walls as they approach the poriferous surface, a character which is well marked in the specimen from the Guadalupe Mountains. As their figures show the apertures of the cells to be separated by considerable intervals, it may be inferred that in G. indica also the walls become heavier near the celluliferous surface. Furthermore, Waagen states that a lamella can usually be seen on the poriferous side of the zoarium in G. indica, which is quite in keeping with the carinated condition of this surface. In G. americana, on the other hand, the plate is seen only near the nonporiferous side.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Family ACTINOTRYPIDÆ Ulrich.

Genus ACTINOTRYPA Ulrich.

ACTINOTRYPA? SERA n. sp.

Pl. XXVIII, figs. 11 to 11b.

This interesting form is known from material so fragmentary that only a small tangential section has been obtained. From this the following description is taken.

The zoarium consists of zoœcia and mesopores. The former are circular and come about four in a linear distance of 2 mm. They stand at intervals from one another varying from one-half to the same as their own diameter, or sometimes even more. The mesopores are very irregular in size and shape. They range in diameter from one-fourth to one-half that of the zoœcia. Apparently they are of a more or less cystose nature, being sometimes angular and sometimes rounded, not infrequently partly one and partly the other. The walls of the zoœcia possess a singular and characteristic structure, seeming to be rather regularly interrupted, or, better, to be subject to more or less regularly distributed areas of densification. These spots in thin section are of about the same intensity of shade as the walls of the mesopores, while the intervening spaces are much lighter, though as a rule they can be distinguished from the calcitic filling of the zoarium and can be traced in their complete circumference. These spots are densifications and not thickenings. They do not form, so far as observed, denticles projecting into the zoœcial interior, though more or less doubtful traces of such structure have been observed in one or two instances. They are of varying length, are not often conspicuously circular, and occur at varying intervals apart.

In almost the majority of cases the mesopore walls when directed so as to intersect that of a zoœcium do not impinge upon a densified segment of the wall but upon one of the translucent segments, and occasionally appear to fall short of contact. This form at first strongly suggests the genus Actinotrypa, and in fact the differences noted may possibly be due to difference in the maturity of the specimens where the section was taken.

Actinotrypa? sera should be readily distinguished from A. peculiaris. One well-marked difference is the continuous, dense, denticulate zoœcial wall of the latter species and the interrupted nondenticulate wall of the former. If it is a real difference (and the denticulate condition begins at an early stage in A. peculiaris) it is such as possibly to demand even a generic separation, but if, as suggested above, this difference is merely a matter of maturity there are still others by which they can be discriminated. The zoœcia in A.? sera are considerably larger and not quite so many occur in a given distance. The mesopores appear to be actually smaller, and relatively the mesopores and interzoœcial distances are distinctly smaller.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2962).

BRACHIOPODA.

Family CRANIIDÆ King.

Genus CRANIA Retzius.

CRANIA sp.

Under this title are subsumed two very imperfect specimens, one from the white limestone of the Capitan, the other from Shumard's "dark limestone." The former is the better preserved, though smaller. It is nearly circular in outline, with a diameter of about 5 mm. or a trifle less. The shape is depressed-conical and the apex is situated about one-third the diameter from the posterior margin. The thick shell is deeply exfoliated, but retains suggestions of rather strong sublamellose concentric markings.

The specimen from the "dark limestone" is somewhat larger, having a diameter of 8 mm. and a height of about 2 mm. The apex appears to be subcentral, but the marginal outline is very indefinite. The surface has been much exfoliated, but was probably marked by faint concentric lines.

Provisionally it seems necessary to refer both specimens to the same species. Their generic position is a little uncertain, but probably lies with Crania. The shell seems to lack the phosphatic constitution which would indicate that it was a discinoid, such as Lingulidiscina; nor do I believe that it is a patelloid gasteropod.

This species is clearly distinct from Shumard's Crania permiana, which is without much doubt a Richthofenia, though it is not without resemblance to Crania modesta White and St. John.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

Family STROPHOMENIDÆ King.

During the Carboniferous epoch one branch of the Strophomenidæ, the Orthotetinæ,a underwent numerous and interesting developments of structure and configuration. The structural peculiarities chiefly relate to the ventral valve. The structures of the dorsal valve, either because they did not undergo corresponding modifications or are less easy of observation, are less distinctive for classification. The ventral valve shows variation in the degree of development of the septum and dental lamellæ, and in their relation to one another. In one type both septum and dental plates are wanting; in another the septum is well developed, but the dental plates are more or less obsolete; in a third both dental plates and septum appear, but the dental plates converge and unite with one another and with the septum in a Y-shaped structure, the dental plates and pseudodeltidium forming a pyramidal chamber on the back of the shell; in a fourth a septum is absent and the dental plates are extended in a discrete condition to the convex shell wall. With these variations in ventral structure are sometimes associated others in the dorsal valve relating to the cardinal process and the development of socket plates. The variation in configuration consists largely in the development in certain groups of species of more or less strong radial plications. There is an almost complete parallelism between structure and configuration, most of the structural types including plicated and unplicated species, the unplicated forms appearing earlier in point of time. Waagen has remarked this circumstance, and commented on it at some length. He says:a

---

aI have changed the form of this name to agree with the corrected spelling of the generic term from which it is derived. I am in doubt about the propriety of retaining this subfamily name. It can hardly be employed so as to have as its central idea the group of species to which the name Orthotetes no longer applies, and I am uncertain how far precedent will warrant shifting the central conception to another group. However, since Orthotetinæ will connote the same genera with either Orthotetes (in its revised sense) or Schuchertella as the central idea, and since there is some doubt as to whether this case has any established precedent, I have consulted my own preferences and retained the familiar term.

aWaagen, W., Salt Range fossils: Main. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 594.

---

A fact that has already occurred three times to our observation, and which can not be passed in silence, is that in several groups of the forms more or less nearly related to Streptorhynchus the geologically younger species attain more or less distinctly radially plicated valves. This peculiarity we had occasion to observe in the genus Streptorhynchus itself, where the form occurring with or above Strept. pelargonatus, viz, Strept. pectinformis and distortus, are strongly radially plicated. Quite the same occurs in the genus Meekella, the Mountain limestone species, M. oliveriana, Vern., being smooth, while the species from the Coal Measures and the Upper Carboniferous limestone, M. striatocostata, Cox, and M. eximia, Eichw., have a strong radial plication. Another instance is the section of the "Camerati," within the genus Derbyia, where the geologically oldest species, Derb. correna, Derb., is not plicated; while the Permian forms, Derb. eusarkos, Abich, and Derb. pereyrina, Ab., are more or less distinctly radially plicated. Lastly, in the section "Septati" of the genus Derbyia a similar peculiarity prevails, though in a much less degree. The geologically older species like Derb. senilis, Phill., Derb. grandis, W., and Derb. regularis, W., are smooth, without a trace of a radial plication; Derb. plicatella, on the contrary, which occurs in the Cephalopoda bed of Jabi, has tolerably strong traces of such a plication. It is now in many instances very highly probable that the plicated forms are the descendants of the smooth ones, but if this be the case it is at the same time very improbable that a character which occurs in absolutely the same manner over the whole world should have been caused by external influences, as climate, food, etc.; there must have existed within these organisms an innate law according to which they were forced to assume with the progress of time, sometimes sooner, sometimes later, a radially plicated shape under most widely different external circumstances.

Many of the types have received names, some of them generic, some subgeneric; and it seems to me that each structural type can appropriately be esteemed of generic rank with the plicated varieties distinguished as subgenera.

Among strophomenids in which the ventral valve has neither dental plates nor septum two divisions have been recognized. For one of these King proposed the name Streptorhynchus, the other was first discriminated and characterized by Waagen, who revived for it the name Orthothetes (properly Orthotetes) Fischer de Waldheim; but I find that Fischer de Waldheim's description of Orthotetes and the type species with which the name must be associated have the characters of that group for which Waagen simultaneously proposed the name Derbya. The term Orthotetes, therefore, passes to the group (or part of it) at present called Derbya, leaving anonymous that now passing under the name of Orthothetes. For this group the name Schuchertella a has been proposed, and Streptorhynchus lens White designated as the type species. Schellwien b employs the term Orthothetes to designate a group of Devonian and early Carboniferous species having two short, strongly diverging dental plates in the ventral valve. The shells which Waagen called Orthothetes, and for which the name Schuchertella has been substituted, Scheliwien refers to Streptorhynchus. I hold to the groups used by Waagen in this instance, though not to his nomenclature, for the term Orthotetes can not be retained, either in the sense in which Waagen or that in which Schellwien employed it.

---

aProc. U. S. Nat. Mus., vol. 27, 1904, p. 734.

bNeues Jahrbuch, Jahrg. 1900, vol. 1, 1900, pp. 6 et seq.

---

In Schuchertella growth is usually regular and symmetrical, the width at the hinge line being equal to or greater than that in front. The two valves meet along a plane which is as a rule nearly perpendicular to the cardinal area. There is no median septum in the ventral valve, and dental plates are also absent, though the edges of the delthyrium may be more or less thickened. The dorsal valve has for the family a rather high area. The cardinal process is bilobed, and in some species of large size. It expands at its base into two winglike supports, which are short and not prolonged so as to surround the muscle scars. This genus begins in the Silurian, but seems to attain its greatest development rather late in Devonian and early in Mississippian time, being more or less completely replaced later on by Derbya and Orthotetes. No North American species of Schuchertella are known in the upper Carboniferous, though one South American form (Streptorhynchus tapalotense) is found at that horizon, and it may be that some of the upper Carboniferous forms referred to Streptorhynchus crenestria really belong here. Waagen describes one species (Orthothetes semiplanus) from the Permian of India. This species Schellwien now places in Streptorhynchus. The Mississippian forms of Schuchertella do not possess dental plates, and the genus can not contain the shells having this structure, for which Schellwien uses the name Orthothetes.

I am not entirely sure that the group of shells with which Waagen and also Hall and Clarke associate the name Orthothetes is distinct from Streptorhynchus. Schellwien, as already noted, throws them together. The differences recognized by Waagen reside chiefly in the dorsal valve. He describes Streptorhynchus as having a large septum supported by two crural plates which partly surround the muscular impressions. In "Orthothetes" he records that the cardinal process is small and not supported by crural plates. He also mentions a low median dorsal septum as usually present.c The dorsal septum is very rare in the forms seen by me. The cardinal process, though often fairly constant in specimens of the same species, yet varies so much, both in size and shape, in different species as to indicate that the value which has been assigned to this structure in discriminating Schuchertella (Orthothetes) and Streptorhynchus has been overrated. The figures given by Hall and Clarke of the cardinal process of Streptorhynchus hallianum show much variability in that structure, both as to size and conformation. I am convinced, however, that more than a single species is represented by these figures. It seems probable that, like other areas of muscular attachment, the cardinal process is subjected in old age to excessive shell secretion, not only increasing its size, but by strengthening its muscular features also modifying its shape, so that the process varies much in both particulars, owing to difference in age. My own observations lead me to doubt if any constant difference exists in this structure available for discriminating Schuchertella from Streptorhynchus. The absence of crural plates in Schuchertella seems to be a constant feature, but I doubt if their presence is so in Streptorhynchus. In Streptorhynchus pelargonatus, the type of the genus, as represented by Davidson's figures or as observed by myself in specimens, the structure of the hinge plate does not differ from that of Schuchertella. The dorsal area of Schuchertella seems to be as a rule higher than in Streptorhynchus, though I doubt if this is constantly the case, or is true in more than a comparatively slight degree. The configuration of the genotype, Streptorhynchus pelargonatus, is peculiar. Its strongly elevated and somewhat twisted ventral valve, narrow hinge line, and curved area of valve junction all distinguish it from the common type of Schuchertella, but these peculiarities of configuration are by no means persistent throughout the forms referred to Streptorhynchus. Among species known to me Streptorhynchus hallianum and Schuchertella tapajotensis do not differ materially in configuration nor in the matter of dorsal area, in both of which they are more like Schuchertella than Streptorhynchus pelargonatus; yet Streptorhynchus hallianum has the crural plates characteristic of Streptorhynchus, while these are absent from Schuchertella tapajotensis.

---

cWaagen, W., Salt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 576.

---

There is another character which I do not recall having seen mentioned in that connection, but which may, if not always an aid in discriminating the two genera, at least serve to show that they are really distinct. Two of the three species of Streptorhynchus noted in this report were attached not by a pedicle issuing from between the two valves, but by cementation of the apex of the ventral valve. The third Guadalupian species was probably attached in the same manner, and some at least of the foreign species, though I have not been able to examine them to ascertain this fact. There is no reason to believe, however, that any species of Schuchertella departed from the normal peduncular attachment. One can not but believe that atrophy of the pedicle accompanied attachment by cementation, entailing with it muscular and other organic modifications, such as must demand a discrimination of Streptorhynchus from Schuchertella on the soft parts, even if the test sometimes fails to show variations to correspond. On the other hand, some of the Derbyas almost certainly practiced cementation, yet they appear to manifest no modifications of structure resulting from it, nor would I advocate separating them on this account from the normal type.

I do not know, therefore, of any characters which can always be relied on to discriminate these two genera, though possibly those already mentioned will be found to serve in a majority of cases. However, the great expansion of the Schuchertella group in the upper Devonian and lower Mississippian, and its practical extinction thereafter, taken with the development of Streptorhynchus in the late Carboniferous and Permian, show to some extent that the stock is not the same, even if the distinctive characters can not yet clearly be designated. In suggesting the term Schuchertella for the species left without a name by the diversion of Orthotetes to the camerate Derbyas, I have not altered the existing group, but only made some necessary changes in the nomenclature.

Streptorhynchus as at present defined includes both plicated and unplicated shells, though the type species is of the latter sort. I think that a valid subgenus might be established for the plicated shells, which Waagen is satisfied with calling the plicati, to distinguish them from the typical series, which he terms the simplices. The fact that the Schuchertella group of forms does not develop a plicated series of species might also be taken as indicating a difference in the origin of the two genera. But two North American species of Streptorhynchus have been described up to the present—Streptorhynchus ulrichi Hall and Clarke and Streptorhynchus williamsi Weller,a both of them from the lower Carboniferous rocks of the Mississippi Valley, though Waagen b believed the genus to be confined to the Permian. The Guadalupian fauna contains three species which seem to belong to this genus, but so far all the American forms are of the unplicated type. Waagen,c however, mentioned having seen specimens from Nebraska in the Royal Paleontological Museum in Munich, in which these septa seem to be absent. I suspect that if really from America, the specimens were Meekellas, which by some accident failed to show their proper structure.

---

a I am not satisfied that these species are not survivors of the group Schuchertella, rather than harbingers of Streptorhynchus.

b Waagen, W., Salt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 3, 1887, p. 576.

c Idem, p. 578.

---

For strophomenids with a median septum in the ventral valve Waagen has proposed the name Derbya. Two variations of this type are found. In one the dental plates appear merely as columnar thickenings terminating below in the hinge teeth, with which they are continuous. They never are extended into distinct places, and since they have about the same degree of development as the corresponding structures in Streptorhynchus, it might with equal truth be said that here also dental septa are absent. These projections vary in degree from being practically absent to appearing as moderately high ridges, and their direction is vertical to the area, or often somewhat diverging. At the apex of the shell the septum connects with the areal wall, but may or may not do so below. The septum and dental ridges come into union only at the apex, where, however, there is sometimes a solid deposit of shelly matter uniting the internal structures with the inclosing walls. To this division Waagen gave the name septati.d In the other, which he calls the camerati, the cardinal teeth are supported by short dental plates, which converge and unite with the septum, forming a small, prismatic chamber in front of the pseudodeltidium. In this group the septum extends no farther than its junction with the dental plates, and only touches the areal wall at the apex. The differences of structure of the camerati and septati is both striking and sustained among the later developed types, but in the Mississippian epoch these septiferous shells show much variability in structural development, some examples of the same series appearing to belong to the camerati and others to the septati. In these forms the dental lamellæ do not unite with the septum for their whole length, and though converging do not completely inclose a chamber where they are free. The duration of their union varies greatly in different individuals. Often, too, the apex of the shell is filled to varying lengths with a solid shelly deposit. In some forms the chamber is rather long, and either open or filled with calcareous matter. In others it is so small and closed with shell that the specimen might without much violence be placed with the septati. The plates in this case of course converge, whereas I have never observed them to have this direction in the real septate group. It may be that the direction of these dental ridges will prove of some importance. The genus Derbya Waagen, as origin ally defined, includes both the septate and camerate types of structure, though the genotype, D. regularis, is one of the septati. The genus Orthotetes Fischer de Waldheim, though misconceived by Waagen to connote the group of shells for which the name Schuchertella is here substituted, really seems to be exactly coextensive with Derbya, though the genotype, O. radiatus, belongs to the camerati. The camerate and septate types of structure, where well differentiated, present differences so striking that I at one time contemplated proposing a new name for the camerate division, not being at that time aware of the true significance of Orthotetes. This idea was later abandoned because of the ambiguous Mississippian forms referred to, but I shall be glad to avail myself of a name already in the literature. By restricting Derbya and Orthotetes to the phase of structure exemplified by the typical species of each, the name Derbya, which conveys a compliment so well deserved, can be retained as well as Orthotetes, which has a long priority; and Since the true Derbyas are much more numerous than the camerati, comparatively few changes in nomenclature will be entailed.

---

d Idem, p. 591.

---

Among both the camerati and the septati Waagen points out species, some of which have a smooth and others a radially plicated surface. Of the plicated Derbyas Waagen mentions only D. plicatella, from the "Cephalopoda bed" of Jabi, and no other species having this character has come to my knowledge. The plications in this species, however, are so faint and irregular as to be far from striking. As instances of plicated shells of the camerate group (Orthotetes) Waagen mentions Streptorhynchus crenistria var. eusarkon Abich and S. peregrinum Abich. After examining specimens of these species, however, Schellwien states that instead of having a single median septum they have two nearly parallel dental plates. He refers them, therefore, to his genus Orthothetina, but if they have a plicated surface, as represented by Abich, they would probably be more correctly placed with Meekella. Shells having radially plicated exterior joined with the internal structure characterizing the genus Orthotetes do occur, however, and Schellwien has recently proposed for them the generic name Geyerella a (type Geyerella gemmellaroi Schell.). The admission of this name with the rank of genus makes an inconsistent and irregular classification if the groups of Streptorhynchus called plicati and simplices are borne in mind.

---

aNeues Jahrbuch, Jahrg. 1900, vol. 1, 1900, pp. 4, 12.

---

The fact that Derbya does not develop a group of species having well-marked radial plications may be considered as having some bearing in estimating the propriety of distinguishing the septati and the camerati as two separate genera.

The genus Derbya is well represented in the Mississippian and Pennsylvanian rocks of North America, but no species of Orthotetes have been recognized with the exception of the ambiguous Mississippian species referred to above. In fact, shells having the distinctive characters of Orthotetes well developed appear not to have come in until after the genus Derbya had been long established. In the fauna described here several species of Derbya are found, but in the Capitan formation Orthotetes occurs in abundance, with the characteristic structure strikingly developed and with a rather peculiar configuration and sculpture. All the American species of Derbya are of the unplicated type. The plicated division of Orthotetes for which Schellwien has recently proposed the name Geyerella, though previously unknown in the western hemisphere, is represented by a species from the Capitan formation for which the name Geyerella americana will hereinafter be proposed.

Both the plicated and unplicated kinds of surface are found among shells having two strong dental plates uniting the anterior and posterior walls. For one of these (the plicated type) the well-known term Meekella has been used, and for the other Schellwien has introduced the name Orthothetina.a There appear to be two groups of unplicated forms having this biseptate structure, one of them, found in the European Devonian, in which the septa are short and diverging, the other widely distributed in the upper Carboniferous and Permian, in which the septa are long and more nearly parallel. It is apparently for the latter group only that the term Orthothetina is employed by Schellwien, and for the septate Devonian forms he uses the name Orthothetes. Species having the structure of Orthothetina are at present not known from North America, though I have an undescribed form from the Permian (?) of Kansas, and a species with converging, closely proximate septa is described below from the Guadalupe Mountains. Nor is the type to which Schellwien misapplies the term Orthothetes known here. The group of species which abounded in this region at the corresponding geologic epocbs is characterized by having the dental plates virtually absent (Schuchertella). Schellwien subsequently seems to have concluded that the more recent septiferous shells should not be separated from the early ones, and accordingly he employs the term "Orthothetes" for the whole, abandoning Orthothetina. But Fischer de Waldheim's name clearly can not be used for shells having extended dental plates but no septum, and accordingly Orthothetina should be recognized. I am, however, by no means content that the early Carboniferous biseptate shells should be referred to the same genus as the late ones, and at the same time it is inappropriate to place them with Schuchertella. Perhaps a new name should be introduced to cover them.

---

a Neues Jahrbuch, Jahrg. 1900, vol. 1, 1900, p. 8. This is the earliest use of this term to which I have found reference, but it is not defined in a formal manner, neither is it accompanied by any indication that this is the first time that it has been employed, nor is any citation of original place of description appended.

---

The genus Meekella is abundant in the upper Carboniferous of the Missisippi Valley, no Mississippian forms being known. Several new species occur in the Guadalupian. Waagen did not find in the Salt Range shells which possessed the structure of Meekella, and having remarked a specimen from Nebraska in which the dental plates were apparently absent and which at the same time possessed the exterior of Meekella striaticostata, he expressed himself as doubtful whether that genus ought not for the most part to be merged in Streptorhynchus. I have not myself observed, nor seen noted by others, any specimen like that mentioned by Waagen, and the occurrence of this type in the Mississippi Valley must be rare. Most American identifications of Meekella striaticostata with little doubt are at least congeneric with that species. The plicated shells without dental lamellæ in the ventral valve, if such really occur here, can not of course be referred to Meekella striaticostata or to the genus Meekella at all.

The Orthotetinæ in their later development show numerous lines of modification. One of the first changes is the loss of an area in the dorsal valve. This structure is well marked in Schuchertella, but seems to be missing in the other Carboniferous divisions, if one may trust the current generic diagnoses. It does occur sporadically in several genera, as I have observed it in Derbya cymbula and Streptorhynchus hallianum. It is probable that this structure is relatively broader in Schuchertella, but that it is never entirely absent.

A very striking character occurring in nearly all the later groups is the development of radial plications. The extension of the ventral valve into exaggerated, high, and distorted shapes is still another feature. Some forms are also cemented to other bodies by the apex during part of their life. This has been observed in Streptorhynchus and in the young of Derbya. Accompanying these other developments is often a shortening of the hinge line, which appears to be a rather constant character in some groups. Perhaps a line of retrograde development is found in Streptorhynchus, in which the absence of dental plates is accounted for by Schellwien as being an atavistic trait.

In his interesting and suggestive paper on the Strophomenidæ a this author repeatedly states that the median septum is a development of the dental plates. To my mind these are entirely independent structures, though all converge and unite at the apex of the ventral valves, where they often merge in a solid shelly mass. I have not seen evidence that the septum was developed from the dental lamellæ any more than the dental lamellæ were developed from the septum. Both structures act with almost entire independence. Hall and Clarke have recently described a species of Derbya in which this is especially marked. Derbya cymbula has as a character a distinct groove down the center of the high pseudodeltidium. This groove is caused by the attachment of the septum to the pseudodeltidium, which continues for a considerable distance, probably 10 mm. or more in some cases. On either side of the septum, yet independently, the rather high dental ridges project. Only at the apex have they any connection. The structures of Geyerella, however, where the long converging dental plates and the septum form a triradiate figure, lends some color to Schellwien's hypothesis, and may indicate that the septum in all types of structure is not formed in the same way. Still, even here the septum may simply unite with the dental plates, instead of being formed from them. Another fact which also might be invoked to support Schellwien's claim is that nowhere (except perhaps in Geyerella and Orthotetes) are the dental lamellæ and septa simultaneously developed in strength. Both structures may be absent, as in Schuchertella and Streptorhynchus, but in other forms either the septum is well developed and distinct dental lamellæ are absent, or the dental ridges are extended into plates and the septum is absent, unless, as above remarked, they unite to form a three-rayed figure. Yet somewhat differently viewed, these facts might better be construed as evidence that the septum and dental plates were supplementary but independent. If the septum in Geyerella is the result of the welded dental plates, and if, on the other hand, the septum in Derbya, as I am inclined to believe, is an independent structure, the course adopted here of distinguishing as independent genera the two original divisions of Derbya is justified, for the structure of Orthotetes is so like that of Geyerella as to leave little doubt that it arose in the same way, and consequently the septum in the two groups, Derbya and Orthotetes, would have had a very different origin. Another circumstance possibly favoring the same discrimination is that Derbya does not devolve into plicated forms, while Orthotetes does so (Geyerella).

---

a Beiträge zur Systematik dec Scrophomeniden des oberen Palæozoicum: Neues Jahrbuch, Jahrg. 1900, vol. 1, 1900, pp. 1-15.

---

The following table shows the structural modifications of the Carboniferous Orthotetinæ and the nomenclature employed:

Classification of Carboniferous Orthotetinæ.

Shells having neither septa nor dental lamellæ in the ventral valve.

Shells having a well-developed median septum in the ventral valve, with the dental lamellæ more or less completely atrophied, and discrete from the septum except at the apex. This group formed the division of Derbya which Waagen designated as the septati. It is also the group to which Derbya regularis Waagen, the type of the genus, belongs. Derbya as originally defined included also another division, called the camerati, to which the genotype of Orthotetes belongs. If the term Derbya can be retained at all, it will only be by separating the camerate and septate divisions as two distinct genera or subgenera and restricting Orthotetes to one and Derbya to the other. This course is here adopted.

Shells having moderately developed dental plates in the ventral valve, which converge and unite, inclosing with the pseudodeltidium a triangular pyramidal chamber. At their union with one another the dental lamellæ unite also with the median septum with which they form a triradiate figure. These shells, together with 3, constitute Derbya Waagen, of which the present group forms the section called camerati.

Shells in which the ventral valve is provided with two more or less long parallel dental plates without a median septum. The plates are prolonged to meet the anterior or convex wall of the shell.

Of these divisions 1a (Streptorhynchus), 3a (Derbya), 4a (Orthotetes), 4b (Geyerella), 5a (Orthothetina), and 5b (Meekella) have been found either in the typical Guadalupian or in beds in adjacent areas supposed to be equivalent. 1b, 2, 3b, and 6 are the only divisions unrepresented in the fauna.

It may well be questioned whether the different groups in the foregoing table which are based on structure are of equal value. Probably they are not so, but the differences seem to be of insufficient importance to have been recorded in the nomenclature. Schellwien apparently recognizes no distinction between groups 1 and 2. Practically all authors place 3 and 4 under the same generic term, but Waagen, at least, recognizes them as subordinate groups. I at one time proposed in manuscript a distinctive name for the camerate division, intending to recognize it as a genus, but subsequently suppressed the name. Some doubt still remains as to the advisability of recognizing the camerati and the septati as independent genera, and I would not venture to carry out this purpose by proposing a new name, though I avail myself of the existence of two appropriate ones already in the literature to tentatively establish this classification. It is true that as originally proposed Orthotetes and Derbya included both shells having a camerate and those having a septate structure; but the type species of Orthotetes belongs to one and that of Derbya to the other division, and it is proposed to restrict each name to the division to which its type species belongs. This course has the additional advantage that it will conserve the terminology as largely as possible in its present form, since the name Derbya has received wide acceptance, and since, comparatively few species of Orthotetes being known, it will continue in use mainly for the group for which it is now in vogue.

While there is room for doubt as to the equality of the different divisions of strophomenids here recognized, there can be little question, I think, as to the irregularity of existing nomenclature for them. For example, Geyerella is distinguished from Orthotetes and Meekella from Orthothetina merely by having a plicated surface; but the plicated group of Streptorhynchus which has the same relation to the simplices is not distinguished as even a subgenus. It is apparent that with a few exceptions each of the structural types recognized above contained species which have simple and those which have plicated shells. A notable exception to this rule is the group Schuchertella, which so far as known is without plicated species. Another instance is the group of Devonian and early Carboniferous shells for which Schellwien erroneously employs the name Orthothetes. As a similar case, may possibly be instanced the genus Derbya in its restricted sense. Only one plicated species of this group is known up to the present, and in it the character is so indistinct and irregular that the question might be raised whether it should be really considered a plicated form. In the case of Geyerella, however, the kindred group Orthotetes has a well-marked plicated division. This difference maybe used as an argument justifying the recognition of Derbya as distinct from Orthotetes.

I am much in doubt as to what taxonomic importance should attach to the plicated shell in this family, but convenience and logic demand that a similar importance should be given to it in each case. From the facts as known it certainly seems lacking in both to retain Meekella and Geyerella as genera and leave the plicated Streptorhynchus as a group of less value than a subgenus. There would probably be few who would advocate giving the plicated groups full generic rank. On the other hand, not many would consent to the reduction of Meekella, for example, to the same rank as,the plicati of Streptorhynchus. An intermediate course, which would perhaps more accurately express the relationship of these shells, would be to recognize the plicated groups of species as subgeneric to the simple ones. The genera which would, in my view, be appropriate for recognition are Streptorhynchus, Schuchertella, Derbya, Orthotetes, and Orthothetina. Possibly an additional group of equal rank should be made of the shells which Schellwien calls Orthothetes in distinction from Orthothetina.

The Carboniferous Orthotetinæ are widely distributed over the earth, and in many places occur in very great abundance. They present variations in certain particulars, such as fineness of liration and height and inclination of the area, to rather wide limits, while maintaining otherwise a rigid adherence to a general type of expression. Here and there a striking and apparently well-marked variety occurs, of which Orthotetes guadalupensis of the present work is an example, but as a rule the different varieties melt into one another so completely that it is extremely difficult to distinguish them, so much so that some of our best known authors have in despair referred all the unplicated types to a single species. This was for the most part before an investigation of internal structures had made much progress, and serves to point another circumstance—that different groups now recognized as genera by reason of structural differences are essentially alike in external expression. In fact, practically the only well-marked external difference which these shells develop consists in whether the surface is simple or plicated, a feature which has been used more for generic than for specific definition. As a result, in comparing the Guadalupian strophomenids with those of other faunas it has been found more practicable to do so rather on the basis of their generic than on that of their specific differentiation, so little individuality being as a rule manifested by the smaller groups.

Considerable variation is shown in the distribution of these higher groups, which can be brought out to better advantage if the comparisons are made in one place; and they can be more briefly and conveniently made in that way than if distributed under the different generic headings. Accordingly, I shall proceed to discuss this matter at the present point.

Strophomenids are represented by numerous structural and specific types in the Salt Range of India. Of Streptorhynchus Waagen recognizes no less than seven species, four of which have unplicated shells and belong to the division which he has called simplices, and three belong to the plicati, a group which is entirely without representation in the Guadalupian fauna. Probably all the Guadalupian species would belong to what Waagen calls the group of Streptorhynchus pelargonatus among the simplices. The rather striking group of S. capuloides has no representatives in the American fauna. The form here described as Orthotetes guadalupensis is strikingly similar in external appearance but entirely unlike in internal structure.

The type of structure which characterizes Waagen's group of the camerati under his genus Derbya is not known in the Productus limestone fauna, but has several fine and characteristic representatives in the Guadalupian. They are here distinguished under the name of Orthotetes. Plicated shells having the camerate structure, for which I have adopted from Schellwien the distinctive term Geyerella, though represented by a characteristic species in the Guadalupian, are not found in the fauna of the Productus limestone. Of the septate division of Derbya, here called Derbya sensu stricto, Waagen recognizes six species. While extremely variable in a few characters, the Derbyas seldom deviate far from a primal typical expression, and as a result the Guadalupian species are in a general way very similar to those of the Productus limestone. In the Capitan formation, however, this group is to a considerable extent replaced by Orthotetes, the most characteristic Derbyas being from lower horizons. They appear to be less numerous, less varied, and less robust than the Indian species. All the American forms are unplicated. Waagen describes one plicated Derbya, but from his figure the plications are so obscure and irregular as to leave one in doubt whether the form really deserves to be so designated.

Under the name of Orthothetes subplanus Waagen describes a species which Schellwien later assigned to Streptorhynchus. As a mere matter of synonymy this form should go to my genus Schuchertella, if it has the characteristics which Waagen ascribes to it, and not to Streptorhynchus. In the Guadalupian I have found no species which it seemed to me could properly be placed in Schuchertella.

Of strophomenoids with two long dental plates, for which the term Meekella has been employed in the case of plicated shells, and Orthothetina in that of unplicated shells, no species are known from the fauna of the Productus limestone. Several well-characterized species of Meekella have been obtained from the Guadalupian, and one doubtfully belonging to Orthothetina. The latter was found in the Capitan formation, but the horizon of the Meekellas is in the Delaware Mountain sandstone.

Another strophomenoid type, very different from the foregoing, which Waagen recognizes in the Productus limestone fauna and doubtfully identifies with Leptæna, is unknown in the Guadalupian.

From this brief survey it appears that the strophomenoids of the Guadalupian on the whole possess but little in common with those of the Salt Range.

In the case of the Himalayan faunas also the resemblance is, so far as known, very slight. In neither of Diener's papers dealing with the "Permo-Carboniferous" fauna of Chitichun No. 1 are any strophomenoids cited, nor in that which describes the Permian fauna of Kumaon and Gurhwal, nor again in the one dealing with the fauna of the Productus shales of the Lissar Valley and of Byans. From Malla Sangcha he has described a species under the title of Orthothetes krafti,a which by a synonymic change should perhaps be written Orthothetina krafti, but the figures show a species so remarkably orthoid in expression as to create a feeling that they really represent a Schizophoria or an Orthotichia, although they do not show the median ventral septum of the latter genus. Nothing similar to this form is known among the Guadalupian strophomenoids. In his paper on anthracolithic fossils of Kashmir and Spiti Diener records Strophomena analoga and Derbya cf. senilis. The former, instead of the characteristic lower Carboniferous species here associated with an upper Carboniferous fauna, I have suggested to be the dorsal valve of a Productus of the P. aagardi group. The other shell is of doubtful generic position.

---

a The generic name being here employed as Schellwien proposed.

---

Writing subsequently on the fauna of the anthracolithic beds of Spiti, Diener figures from the lower beds a shell which also has no analogous species in the American fauna. He identifies it as "Derbyia cf. senilis," but it seems to be quite different from the species cited under that title in the previous paper. Not long ago Professor Schuchert called my attention to this form, which, he wrote, looked to him like an Atrypa, a resemblance that to me also is singularly close. The form in question occurs in beds which contain a spiriferoid of the genus Syringothyris (described as Spirifer curzoni), a fact which would indicate an early fauna of the Carboniferous, and although Atrypa is not, I believe, definitely known from Carboniferous strata, it is possible that still earlier faunas occur at the locality. Diener mentions other shells belonging to "Derbyia," some of which he thinks resemble D. crassa of our American Pennsylvanian. From the upper horizon he obtained one of the plicated species of Streptorhynchus, cited as Streptorhynchus cf. pectiniformis, a type which is notable by its absence from the Guadalupian. It is perhaps this same species which as a small plicated and striated dorsal valve Davidson figures from Kashmir under the title Streptorhynchus ? sp. An unplicated shell is figured as Streptorhynchus crenistria.

From Turkestan Romanowsky figures one of the ordinary Orthotetinæ which might belong to one of several genera, so far as can be determined from the text and figures. It is identified as Streptorhynchus crenistria.

From China, in the Lo Ping fauna, Kayser cites only Streptorhynchus crenistria var. senile Phillips and Meekella striaticostata Cox?. It is a rather singular form, or group of forms, which is figured as Streptorhynchus crenistria var. senile, and a reexamination of Kayser's material by Schellwien and by Fliegel has resulted in subdividing it into several genera and species. Fliegel recognizes no less than five, viz, Streptorhynchus kayseri, S. subpelargonatum, Orthothetes [Orthothetina] circularis, Orthothetes [Orthothetina] kayseri, and Derbya sp. Streptorhynchus kayseri, from its large size and configuration, is rather distinct from the Guadalupian representatives of the genus. The species S. subpelargonatum, which is not figured, would perhaps have more in common with them. The single specimen of Orthothetina from the Guadalupian is too imperfectly known to stand for much in the way of comparison, but so far as one can tell it does not differ widely from O. circularis and O. kayseri, which are closely related to one another. The undetermined species of Derbya found in the Lo Ping fauna does not form a practicable basis for comparison with congeneric types in the Guadalupian. Kayser's Meekella striaticostata is almost certainly not our common Pennsylvanian species, and it may be doubted whether it is a Meekella at all, instead of one of the other types having a similarly plicated exterior but different internal structure. Indeed, Fliegel states that the plication in this specimen is by no means as distinct as represented in Kayser's figure and that it may well be a representative of his species Orthothetina kayseri.

In the various imperfectly known faunas described by Loczy from the Carboniferous of China, strophomenoids are cited in but two instances, and in each case they are identified as Orthothetes crenistria. One of the localities furnishing this species was Tengtjantsching and the other Youngtschangfu. Loczy appears to have followed the classification introduced by Waagen, and if so his shell would probably be a Schuchertella, a group which I have not found represented in the Guadalupian fauna.

From the Carboniferous of Padang but two strophomenoid species are at present known, one of which Roemer cited as Streptorhynchus crenistria var. senile. This has later been determined by Fliegel to be a new species, and has been called by him Orthothetes [Orthothetina] politus. The other is briefly characterized under the title Orthothetes [Orthothetina] sp. The Guadalupian representative of this group is too poor to permit a comparison.

From the so-called Permian of Timor and Rotti, in the Indian Archipelago, Rothpletz cites two strophomenoids which he calls Streptorhynchus cf. crenistria Phillips and Streptorhynchus beyrichi n. sp. I am unable to determine whether the generic name Streptorhynchus is here used in the general or the restricted sense. The fragmentary specimen of Streptorhynchus beyrichi, which alone is figured, is of the ordinary type.

Beyrich had previously identified and figured these species as Streptorhynchus crertistria? and Streptorhynchus radialis. Externally they belong to the usual unplicated type, and there is no clue as to their internal structures.

Martin cites from Timor as Streptorhynchus cf. pectiniformis Davidson a small, rather strongly plicated species, presumably a Meekella, but possibly a true Streptorhynchus. In the former case it appears to be of the usual type and is related to several Guadalupian species. If a Streptorhynchus, the Guadalupian contains nothing like it.

The Strophomenidæ of New South Wales as described by De Koninck comprise only two species—Strophomena analoga and Orthotetes crenistria. The horizon of both species seems to be in the lower part of the Carboniferous section, which is probably much older than the Guadalupian.

The Strophomenidæ of the "Permo-Carboniferous" of Queensland and New Guinea comprise, according to Etheridge, but two species—Strophomena rhomboidalis var. analoga and Derbya senilis. In several other instances where Leptæna rhomboidalis was cited from some of the higher beds of the Carboniferous, I have thought there was some reason to believe that the identification was based on the dorsal valve of some productoid, such as Productus aagardi, or P. waagenianus of the present memoir. Some at least of Etheridge's figures seem to represent real Leptænas, and I can not avoid the suspicion, since that genus is restricted in its upward range in the United States, and in Europe so far as I am aware, to the lower portion of the Mississippian or sub-Carboniferous, that these Australian formations (the Star and the Gympie, but not the Bowen River) are not "Permo-Carboniferous," but much older. The suspicion is based not on this one species alone, but on the generality of them.

If we may depend on the figures, which are freely given, Derbya senilis is not a Derbya, but a Streptorhynchus. It is represented as without a septum or dental plates and with two well-developed socket plates. It is, however, evidently not closely related to any of the Guadalupian species of Streptorhynchus.

Etheridge, senior, described Streptorhynchus davidsoni from Queensland, and also cited from a different locality Strophomena rhomboidalis var. analoga. The Streptorhynchus is certainly not one of the common strophomenoids if reliance may be placed on the figures, and its generic relations appear to be doubtful.

As to the range of this group in the Russian section, I can not hope to have gained complete data, yet from the works which have come under my observation probably the most important facts can be gathered. From these the Carboniferous strophomenoids seem to have reached their maximum development in the "Upper Carboniferous" or the Gschelian. From the Productus giganteus zone I have seen cited only Streptorhynchus radialis and S. crenistria, the term Streptorhynchus here probably being used in its comprehensive sense, so that one might expect almost any of the common types of Schuchertella, Derbya, Orthotetes, etc.

From the Spirifer mosquensis zone I have found cited Orthotetes crenistria, Derbya grandis, Derbya sp., Meekella eximia, and Meekella sp. It is this fauna for the most part which is treated by Trautschold in his work on the fauna of Mjatschkowa. In this work Trautschold discusses Orthis crenistria, Orthis senilis, and Orthis eximia, which belong to the Strophomenidæ, besides some true representatives of the Orthidæ. The figure of O. crenistria given by Trautschold clearly belongs to Waagen's genus Derbya, but whether it shows the septate structure (Derbya sensu stricto) or the camerate (Orthotetes) can not be determined with absolute certainty. It appears to belong to the camerati, however, and may be a representative of Fischer de Waldheim's species Orthotetes radiatus, the genotype of Orthotetes. If so, it is not like the Guadalupian representatives of Orthotetes, which, with the exception of the doubtful form from the base of the section, are distinguished by having tall, conical ventral valves. Much more similar in this respect is the apparently smooth form which Trautschold figures as Orthis senilis. It is impossible to tell in what division to place this shell. Orthis [Meekella] eximia is one of the ordinary types of Meekella, and does not depart so widely from the American Pennsylvanian or Guadalupian species.

From the Gschelian I have found listed a large number of strophomenoids, to wit: Orthotetes crenistria, Derbya senilis, Meekella eximia, Meekella striaticostata, and Meekella cf. eximiiformis. Tschernyschew in his monograph on the Brachiopoda of this fauna distinguishes a much greater variety. In the genus Streptorhynchus he identifies S. pelargonatum, S. hallianum, and S. aff. tapajotense. S. hallianum is one of the plicated group of Streptorhynchus, a type not yet known from the Guadalupian while S. pelargortatum and S. tapajotense, although I believe them to be distinct specifically from the Guadalupian species, are yet more nearly of the same general character. Of Derbya this author identifies two of Waagen's Indian species, D. regularis and D. grandis, together with D. crassa, our common American Pennsylvanian form. I notice that he includes Orthis crenistria Trautschold, to which I have referred above, in the synonymy of D. regularis. These species are of the same general type as the Guadalupian ones, and, in fact, more or less similar species are found at different horizons the world over. Of Meekella Tschernyschew distinguishes no less than seven species; but I am not sure that he does not include among them some which would more properly be placed with Orthothetina. Meekella ufensis, M. baschkirica, and M. uncitoides have shells so slightly folded that one can in the figures detect it with difficulty, if at all. If we except these from Meekella, no very marked differences can be pointed out between the Gschelian and the Guadalupian representatives of the genus. Perhaps the Guadalupian species M. difficulis, with its angular plications and almost obsolete liration, might be cited as an exception to this statement, which is, however, further borne out by M. attenuata, in which the plications are inclined to be faint, and by Orthothetina sp., which has a smooth shell, but with the internal structures of Meekella. Tschernyschew also describes a species of Orthothetes (O. simensis) from these beds, the generic term being employed in the sense used by Waagen, O. simensis is represented as possessing faint radial plications near the margin, a feature not before noticed in this group, so far as I am aware. I find myself a little disposed to follow Schellwien in referring the Permian shells which have this structure to Streptorhynchus.

From this or possibly lower horizons De Verneuil has described and figured strophomenoid shells, which he calls Orthis arachnoidea, O. eximia, and O. olivieriana. The first mentioned may possibly be a Schuchertella; the second is Meekella eximia of Tschernyschew's and other reports; and the last-named species one would not hesitate to call, from De Verneuil's figures, a characteristic representative of Orthothetina.

In view of Tschernyschew's extensive memoir on the brachiopods of this fauna, the other references to Gschelian strophomenoids can well be passed over, especially as they are for the most part citations in lists, without descriptions or illustrations. From the Artinsk Stuckenberg cites Streptorhynchus crenistria and from the closely related Kungurstufe Streptorhynchus crenistria and Meekella eximia. Krotow also records only Streptorhynchus crenistria from the Artinsk. Tschernyschew is authority for the citation of Streptorhynchus pelargonatum, and he has figured specimens from this horizon. We thus have the genera Streptorhynchus and Meekella in the Artinsk, and probably Derbya or Orthotetes, though it is impossible to tell what is intended by the name Streptorhynchus crenistria.

From the Permian Netschajew has figured a small, imperfectly preserved shell, which he calls Streptorhynchus cf. pelargonatum. One can tell little about this form from the poor figures, and the text is in Russian.

The sudden suppression of the strophomenoids, along with most of the other brachiopodous groups, before the commencement of the Artinsk, leaves the greatest show of resemblance between the Gschelian faunas of the Russian section and the Guadalupian. This resemblance is in some respects rather close. In both is found a considerable differentiation of the genus Meekella. In both the type of Orthothetina is probably represented, in the Guadalupian by an undetermined species, in the Russian beds by O. olivieriana, and perhaps by some of Tschernyschew's species of Meekella. Derbya, in the sense that that term is here employed, is present in both faunas and of the same general type. Orthotetes is represented in the Russian section by at least one species, O. radiatus. In the Guadalupian fauna camerate shells are also developed, the nearest allies to the Russian form being in the lower part of the section. The Russian faunas certainly contain, so far as known, nothing analogous to the conical, highly characterized Guadalupian species, especially to Orthotetes guadalupensis and its allies, nor do they contain any plicated examples of this type (Geyerella), a representative of which is known in the Capitan fauna. On the other hand, the Russian faunas contain both plicated and unplicated types of Streptorhynchus, while in the Guadalupian only the plicated type is known, though it occurs at widely different horizons. Lastly, Tschernyschew describes one species of "Orthothetes," a type which, if the same as Schuchertella, as Waagen appeared to think, has no Guadalupian representatives. On closer inspection of these data it appears, in a general way, that the resemblances are strongest between the Delaware Mountain fauna and the Gschelian stage and the differences strongest between the Capitan fauna and the Gschelian, though this may not be uniformly the case. In this connection one may recall that Orthotetes, of the guadalupensis group, and Geyerella are found only in the Capitan division, not in the Delaware Mountain fauna nor in the Gschelian, while the Meekellas are in the Guadalupian section confined to the Delaware Mountains, nor do they extend above the Gschelian in Russia. On the other hand, we have in the Guadalupian section Streptorhynchus of the same general type in both the Delaware Mountain and Capitan formations, and, likewise, in Russia Streptorhynchus occurs in the Artinskian and Permian, as well as in the Gschelian.

Enderle found only two types representing this group in the Carboniferous fauna of Balia Maaden, in Asia Minor. One of these he cites, without figures, as Orthothetes sp., comparing it with O. subplanus of the Salt Range of India; the other he calls Streptorhynchus cf. pelargonatum. The shell which he figures under this name is distinct from any of the corresponding species in the Guadalupian.

In the Armenian fauna described by Abich from the vicinity of Djoulfa, members of this group seem to occur in considerable abundance, and Abich recognizes no less than twelve types, which he describes as varieties of Streptorhynchus crenistria or of Streptorhynchus peregrinum, a new species. Streptorhynchus is here employed in a general sense, and some of these types are now known to belong to other genera. Waagen refers three of them (Streptorhynchus crenistria var. eusarkon, Streptorhynchus crenistria var. incurvum, and Streptorhynchus peregrinum) to his genus Derbya, supposing them to be representatives of the camerate division thereof,a but Schellwien has shown that in Streptorhynchus crenistria var. eusarkon the two dental plates do not unite with the median septum, but remain parallel and distinct, as in Meekella. He refers this species, therefore, to his genus Orthothetina, although Abich represents it as having a plicated shell. Arthaber later redescribed this Armenian fauna, recognizing in his work three species of Orthothetes (Orthothetina), O. armeniacus n. sp., O. eusarkos Abich, and O. peregrinus Abich, in the synonymy of which most of Abich's names appear. Orthothetina seems to have reached a special degree of development in this Armenian fauna, for elsewhere it is as a rule rather rare. The single Guadalupian specimen which can be referred to this genus unfortunately does not permit comparisons with the Armenian forms. While rich in Orthothetinas, the Armenian fauna contrasts with the Guadalupian in lacking, so far as known, the more varied strophomenoid differentiation. In association with Arthaber, Frech studied the lower faunas of the Paleozoic section, among which he cites Orthothetes crenistria and O. crenistria var. kellii from the earlier Carboniferous deposits at Arpatschai, but this fauna does not concern us.

---

aWaagen, W., Salt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 592.

---

I am unfortunate in being unable to consult that portion of Gemmellaro's work in which he treats of the strophomenoids of the Fusulina limestone of Palermo, for one might expect from the resemblance shown by other groups that comparisons of this one would prove of considerable interest.

Schellwien recognizes three species of strophomenoids in his paper on the fauna of the Carnic Fusulina limestone. They are identified as Orthothetes semiplanus, Derbya waageni, and Derbya expansa. To the first the Guadalupian appears to contain no analogous species, but the two Derbyas resemble the Guadalupian representatives of the genus. Gortani cites from the same region Orthothetes crenistria, O. crenistria var. senilis, Orthothetes expansus?, Streptorhynchus semiplanus, Derbya grandis, Derbya altistriata, and Meekella vinassai. His figures of the last-named species, which indicate that the original specimens were poor, do not represent the surface as plicated, and consequently the species would appear to belong to Orthothetina, rather than Meekella. The Derbyas appear to be of the ordinary type and not materially different from the Guadalupian Derbyas, I judge that Streptorhynchus semiplanus, which Schellwien first described as an Orthothetes (in Waagen's sense), is really not a Streptorhynchus, strictly speaking, but a Schuchertella and consequently a member of a group not found in the Guadalupian. Non-Guadalupian also are probably the species of Orthothetes, for I believe Gortani is using this term as Schellwien interprets it, for species having two short diverging dental plates but no septum.

In his paper on the fauna of the Trogkofelschichten Schellwien finds strophomenoid species belonging to the genera Streptorhynchus, Meekella, and Geyerella. It is surprising to note the absence of both groups of Waagen's genus Derbya in this fauna, for these shells are usually abundant. Schellwien cites but two species of Streptorhynchus—S. pelargonatum and S. cf. operculatum. Both belong to the unplicated type of the genus, and are not very unlike the Guadalupian forms, though I have not recognized among the latter representatives of the group of S. capuloides, to which S. operculatum belongs. Of the genus Meekella the Trogkofelschichten contain five species, agreeing in this respect with the Delaware Mountain formation of the Guadalupian, but contrasting with the Indian Salt Range fauna and the Guadalupian Capitan. The Meekellas are said to constitute an important element in the fauna of the Trogkofelschichten, and they form a rather well-marked group, to which those of the Guadalupian are not closely allied. Schellwien's species in general are distinguished by being large, with numerous small and more or less indistinct plications. Strongly in contrast to them is such a shell as Meekella difficilis of the present work. Of Geyerella Schellwien cites but a single species. He had already described a species from Sicily, which stands as the type of the genus, and I infer that there are others from the same area. Thus the Guadalupian fauna shares with that of Palermo, of Trogkofel, and of Auernig the only known representation of the genus. On the whole, though showing some notable points of difference, especially in the matter of Derbya and Orthotetes, the American fauna appears in its strophomenoid representation rather closely allied with the Alpine one. In the Upper Carboniferous of the same area Schellwien describes two new species of Derbya, of the depressed-convex, regular type, and identifies Orthothetes subplanus, a species which he subsequently assigned to the genus Streptorhynchus. The absence of such groups as Orthotetes sensu stricto, Meekella, Geyerella, Orthothetina, etc., is noteworthy.

The fauna described by De Koninck from Bleiberg, in the Carinthian Alps, little concerns the present discussion, as it is a different and older fauna.

Geinitz, in his monograph on the Dyas, recognizes only one species belonging to the Strophomenidæ, the well-known Streptorhynchus pelargonatum, While some of the Guadalupian types of Streptorhynchus probably belong to the same group with S. pelargonatum, none can rightly be placed in the same species. Of course the American fauna has a much more varied strophomenoid representation than that of the Dyas.

The Carboniferous faunas of Spitzbergen and Nova Zembla concern us little. The latter contains but two forms demanding attention. Strophomena depressa may be supposed to be the same as Leptæna rhomboidalis, but it looks like a Productus. Orthis eximiiformis, cited also from the Russian section, is probably a Meekella. It is founded on a fragmentary dorsal valve.

Toula cites from the south point of Spitzbergen a shell which his figures show to be without much doubt a Derbya of the common type. This form is identified as Streptorhynchus crenistria. As S. crenistria var. macrocardinalis he describes a species from the cape between the two arms of North Fjord which is probably of a different genus, as it apparently has no septum, and seems to be without Guadalupian allies. Lundgren cites Streptorhynchus pelargonatus from the Permian of Spitzbergen. Altogether the Spitzbergen faunas are rather indifferent in their relationship to the Guadalupian.

The Permian of England, like that of Germany, seems to contain but a single representative of this group, and it is the same species, Streptorhynchus pelargonatum, so that no additional comments are necessary.

The fauna which Stache describes from the West Sahara appears to be much older than the Guadalupian, and so far as known contains but two species of the Strophomenidæ, identified as Streptorhynchus crenistria and S. pusillus. "Streptorhynchus" seems here to be employed in the general rather than the restricted sense. If, as seems probable, the African forms belong to Schuchertella, there is nothing in the Guadalupian which can be compared with them.

In the Brazilian fauna, which Derby describes with such ability, he recognizes in all three strophomenoid species, which he calls Streptorhynchus correanum, Streptorhynchus hallianum, and Streptorhynchus tapajotense. Waagen has already discussed these species, calling attention to the fact that the first possesses the structure of his camerate group of Derbyas, the second belongs to the plicated group of Streptorhynchus, sensu stricto, while the third is one of the septate Derbyas. In this fauna we note that Meekella, Geyerella, Orthothetina, and the group of Streptorhynchus called simplices are wanting. These are found in the Guadalupian, which lacks, on the other hand, the plicated Streptorhynchus. Both faunas possess similar types of Derbya and Orthotetes, but the striking group of O. guadalupensis is peculiar to the North American.

The upper Carboniferous strophomenoids of the Mississippi Valley and Appalachian region, so far as known, appear to be divided between the genera Meekella and Derbya. Many of these shells were originally described as belonging to Streptorhynchus, but that genus is apparently not known in this fauna. Waagen, it is true, mentions having seen in Munich a specimen from Nebraska identified as Meekella striaticostata, in which the characteristic dental plates were not developed. But I can not help thinking either that there was a mistake in the locality, or that the dental plates had somehow been destroyed. Otherwise the shell noted by Waagen would be one of the plicated group of Streptorhynchus, a type otherwise unknown from North America. I should remark, however, that from California I have seen a plicated shell which seemed to belong to Streptorhynchus. It is an internal mold, somewhat weathered, and the impression of the plates may have been lost. Mention may also be made of an undescribed species of Orthothetina from the upper Carboniferous of the Mississippi Valley and Wyoming. Of Derbya perhaps eight species in all are known from the upper Carboniferous of central and eastern United States, all of which belong to the septate division, or to Derbya proper. Some of them, as well as the single species of Meekella, at present recognized from this area, have Guadalupian representatives more or less closely related. Orthotetes, Geyerella, and Streptorhynchus, however, are found in the Guadalupian, but not in faunas with typical Pennsylvanian facies. This statement may, however, be only partially true of Streptorhynchus, and is made with the qualification rendered necessary by Waagen's observation noted above.

Genus STREPTORHYNCHUS King.

Schellwien employs the term Orthothetes for a group of shells for which Waagen did not provide—those having two strong dental plates without a median septum—while the group for which Waagen used Orthothetes he unites with the genus under discussion. Both Waagen and Schellwien are in error in their employment of the term Orthothetes, as I have elsewhere attempted to show; but it is surprising that these two authors, using the same species as a type, should have arrived at two different sorts of structure for the genus. Waagen's conception of Streptorhynchus crenistria was undoubtedly influenced by if not derived from Davidson, who probably knew what the characters of the species really were; but Schellwien also invokes Davidson as representing Streptorhynchus crenistria with two dental plates, just as Waagen implies that he represents it without any. I must confess that Davidson's fig. 5., Pl. XXVI, which represents the interior of a ventral valve, appears to me to be without the strong dental plates claimed by Schellwien. As for the small dental or rostral plates which Davidson, as Schellwien points out, certainly does describe as strengthening the hinge teeth, these are as likely to be ridges, like those of Streptorhynchus, as plates such as Schellwien represents in his figures. Indeed, Davidson uses almost the same words, "dental ridge or plate,"a in describing the interior of Streptorhynchus as based on S. pelargonatus that he does in the case of Schuchertella (Streptorhynchus) crenistria.

---

a Mon. Permian Brachiopoda, 1858-1863, p. 30.

---

As elsewhere set forth, I have found it necessary to divert the name Orthotetes from the group for which Waagen used it, and to substitute the new term Schuchertella for it, naming for the type of Schuchertella the well-known Streptorhynchus lens from the Kinderhook of Missouri. This species is characteristic of a certain group of species found in abundance in our American lower Mississippian rocks. The dental plates in these shells are represented only by a thickening, sometimes very slight, of the edges of the delthyrium, and the dorsal valve contains no socket plates except very short, stout, curved ridges. If Streptorhynchus crenistria shows these characters, as Davidson's figures and description indicate, it belongs to Schuchertella. If it shows those represented by Schellwien's figures it does not, but can probably be referred to Orthothetina, which I would provisionally extend to include these forms. Schuchertella is exactly equivalent to the Orthothetes of Waagen and of Hall and Clarke, and includes shells which Schellwien merges with Streptorhynchus. With the procedure of Schellwien I can not agree. Schuchertella was developed chiefly in early Carboniferous time; Streptorhynchus near the end of it. In Schuchertella the ventral valve is mostly low, regular, and unattached; in Streptorhynchus it is mostly high, distorted, and probably attached. In Schuchertella the hinge line is usually long; in Streptorhynchus it is usually short.a In Schuchertella the dorsal valve has a distinct area but no socket plates; in Streptorhynchus the dorsal valve has no area,b rather strong socket plates, and a more deeply grooved and a somewhat differently shaped cardinal process. While it is doubtless true, as Schellwien claims, that considerable variation is manifested in these particulars, I think that the differences will be found numerous enough and constant enough to demand the recognition of both groups.

---

aThat is, shorter than the width in front.

bAccording to Hall and Clarke, who describe it as linear. Davidson (Mon. Permian Brachiopoda, 1858-1863, p. 30) says that there is a small, narrow rudimentary area. It is probably not absent, but as a rule narrower than in Schuchertella.

---

Waagen recognizes two divisions of this genus in the Salt Range, one of which he calls the "simplices" and the other the "plicati." The former, as the name indicates, embraces species having a smooth and the latter those having a plicated shell. In the Salt Range, according to this author, the simplices are restricted to the lower and middle divisions of the Productus limestone, while the plicati occur chiefly in the upper division. In North America only two species of Streptorhynchus are known—S. ulrichi and S. williamsi. To these may be added S. hallianum from Brazil, and three species, to be described later, from the Guadalupe Mountains. All of these species belong to Waagen's division of the simplices, the plicati not being known in either of the Americas, so far as I am aware. Waagen,c however, states that he has seen a specimen, apparently from Nebraska, in the Royal Paleontological Museum of Munich, which has the exterior of Meekella striaticostata, but which appears to lack in the ventral valve the septa characteristic of Meekella. If this occurrence is authentic, the plicated group would appear to be represented in the "Coal Measures" of the Mississippi Valley. This one would hardly expect from its occurrence at the top of the Salt Range section associated with very different genera and species. Furthermore, so many specimens of Meekella striaticostata from the Mississippi Valley have been observed by other paleontologists, and I have myself seen so many, that while possible, it seems not altogether probable that this representative of what must be a very rare type should find its way into a foreign collection, none being known in this country. Therefore Waagen's suspicion that the genus Meekella ought to be in large part merged with Streptorhynchus is unfounded so far as the citations in American literature and the occurrence in American formations are concerned. The range of this genus in the Carboniferous of North America is somewhat peculiar. S. williamsi and S. ulrichi are both from the Mississippian series of the Mississippi Valley, and no representatives of the genus are known from the Pennsylvanian or so-called Permian beds of the same area. Its reappearance in the Guadalupian fauna is therefore of some interest.

---

cWaagen, W., Salt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 3887. p. 378.

---

Still further novelty would be added to the range of the genus if the plicated type supposed to be a later development of the simple one, were really, as Waagen claims, found in the "Coal Measures." The peculiarities of range in American rocks which might be taken as evidence against referring S. gregarium to Streptorhynchus are equally or even more cogent against assigning it to Schuchertella.

This genus is sparingly represented in the Guadalupian fauna. Waagen found seven species in India and Schellwien two in the Carnic Alps. The fineness of the radiating liræ of the Guadalupian species (except S. pygmæum), a character which it shares with representatives of related genera associated with it, invites remark. In this particular, though not in form, comparison can be made with Waagen's Streptorhynchus operculatus.

If one were asked to divide the Guadalupian species of Streptorhynchus into subordinate groups he would probably at once place S. gregarium and S. perattenuatum in one subdivision, since they have lofty ventral valves and fine sculpture, and into another S. pygmæum, which has a low ventral valve and relatively coarse liration, With S. pygmæum appears to go the imperfectly known species Streptorhynchus sp. a. Such at the present time appears to be the natural grouping of these species.

STREPTORHYNCHUS GREGARIUM n. sp.

Pl. XI, figs. 3 to 7.

This form is fairly abundant in the Guadalupe Mountains, and can not be passed over unnoticed, though I have not been fortunate in obtaining complete specimens on which to base descriptions and figures. Much of my material came from a single fragment of limestone, of rather small size, in which the different specimens were crowded together, and their mutual attitude in some cases suggests that they may have been attached one to another. The matrix, in this case a gray, compact limestone, adheres closely to the fossils, and is, moreover, both hard to the tool and difficult to distinguish from the shell substance. The following description is drawn up from a number of specimens, no one of which shows all of the characters.

Shell small. Ventral valve high, conical. Growth usually extremely irregular and contorted. Area flat transversely, more or less concave longitudinally, and inclined backward at an angle of 135° or less to the plane of the edge; more or less poorly defined from the convex portion of the shell. Pseudodeltidium broad and divided into three parts, the central of which is strongly convex and elevated. The lateral portions, which are more nearly on a plane with the rest of the area, represent the internal ridges supporting the hinge teeth. These do not run out into septiform plates, and in one specimen appear to be tubular.

The dorsal valve is moderately or decidedly convex. The shape is subcircular, but varies greatly to correspond with the irregular form of the ventral. The hinge-line is almost always shorter than the shell below. The beak is small, but the umbo is sometimes tumid. There is frequently a slight fold in this valve corresponding to a faint sinus in the other, chiefly noticeable on the anterior margin, where they produce a distinct sinuosity.

The surface is marked by very fine radiating liræ, which are thin and sharp, though but little elevated, and rather distant. They come about 18 or 19 in the space of 5 mm.

Internally the ventral valve is without septiform plates of any sort, though the hinge teeth are supported by strong ridges. The interior of the dorsal valve is imperfectly known to me. There is a distinct median septum, but socket plates have not been observed. The generic position of this form, therefore, is not conclusively ascertained, but the possibilities are reduced among known genera to Streptorhynchus and Schuchertella. To a certain extent in the presence of a dorsal septum, but much more in the absence of socket plates, if they are indeed absent, the affinities appear to be with Schuchertella, but the general expression is so much more that of Streptorhynchus that it has been assigned to the latter genus.

There is scarcely a feature in which this species does not show considerable variation, and I have been forced to assign to it wider specific limits than might be wished. I feel that more perfect material might permit me to make assurance of the presence or absence of characters where now I am in doubt. In general appearance this form much resembles that which I have described as "Orthotetes distorta," but the presence in the latter of dental plates and the extravagant median septum remove the two forms not only into different species but into different groups.

Associated with the form whose more important characters have just been enumerated is another, represented by a single incomplete ventral valve, which should probably be distinguished as a distinct variety. It differs from the other specimens chiefly in being depressed and more rapidly expanding. The growth is irregular and asymmetrical. The width, which was probably greatest at or near the hinge line, must have approached 20 mm; the height probably did not exceed 5 mm. The surface ornamentation consists of regular, sharp, low, radiating striæ, about 22 in 5 mm.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

STREPTORHYNCHUS PYGMÆUM n. sp.

Pl. XXX, figs. 3 to 6b.

This species is fairly abundant at station 3763, but while a considerable number of dorsal valves have been found only three ventrals have yet come to hand. The size is very small, a large dorsal valve having a width of about 7 mm. and a length of 6 mm. The ventral valve is high, erect, and subconical. The aperture contracts posteriorly, is as a rule slightly wider than long, and, owing to a tendency in the lateral outlines to make straight lines near the hinge, is somewhat imperfectly polygonal. The area is high and flat and makes an angle of about 90° with the aperture. Its width is slightly less than that of the shell in front, and it is defined by an angle from the curved antero-lateral wall. The large and strongly elevated pseudodeltidium occupies about one-third of the entire area. Two of the three ventral valves are attached by cementation over a considerable part of the antero-lateral surface and the remaining individual shows a large scar of attachment over this area. This seems to demonstrate a persistent habit, which probably accounts for the relatively large number of free dorsal valves which have been found. For the same reason the surface characters are obscured over most of the ornamented area, and are best studied in the accompanying valve. On the interior the progression of the hinge teeth is marked by two prominent ridges, which are not, however, prolonged as plates, while anything in the nature of a median septum is quite absent.

The dorsal valve is very small and subcircular in outline, the growth more or less irregular and asymmetrical, the convexity moderate to great. The hinge line is shorter than the width below. The surface is marked by comparatively strong, thin, high, more or less irregular, alternating, spaced liræ, crossed by strong but irregular crenulations. Lamellar varices of growth seem to be a common feature of the external surface. The hinge plate and cardinal process are large and massive, projecting from the shell at nearly right angles to the plane of its edges. A low median septum is usually more or less plainly developed. The high convexity of this shell and its relatively massive thickness would indicate that these are mature and not ungrown specimens, an inference which is supported by the general uniformity in size of the specimens collected. With the thickness of the shell appears to be connected the more obvious development of the septum, since it seems to be one of the results of calcareous deposition produced by old age.

I have been in some uncertainty whether to place this form with Streptorhynchus or with Schuchertella, which is very similar in general structure. Schuchertella in this continent is practically confined to the earliest beds of the Carboniferous, types with a well-developed median septum soon succeeding it. Streptorhynchus, on the other hand, seems to be mostly a late Carboniferous development, and in its distribution to have been confined largely to European and Asiatic waters, Nevertheless, certain forms, presumably congeneric with Schuchertella, have, under the name of "Orthothetes," been cited from late Carboniferous strata of India, etc., while in the United States the only accepted species of Streptorhynchus are found in Mississippian horizons.

Aside from stratigraphic and geographic occurrence, the main differences between the two genera seem to be (1) that of configuration, the ventral valve of Streptorhynchus being high, contorted, and contracted at the hinge, instead of low, regular, and with quadrate cardinal angles; and (2) that of structure, the cardinal process of Streptorhynchus being perhaps somewhat differently shaped and the socket plates, according to Waagen, prolonged so as partly to surround the muscular area of the dorsal valve. Different species, however, appear to have varied essentially in the structure of the cardinal process, while figures of interiors of Streptorhynchus seldom show much difference from Schuchertella in the prolongation of the socket plates. In the case of the present species I can distinguish no difference in the structure of the hinge plate from that found in typical examples of Schuchertella, while the configuration of the shell is distinctly in better agreement with that of Streptorhynchus, but a more essential difference seems to subsist in the fact that Schuchertella, like Derbya and Orthotetes, was attached by a pedicle issuing from beneath the lower edge of the pseudodeltidium, while the present species was almost certainly attached by cementation upon the other side of the valve. Whether this very marked difference from Schuchertella is one of agreement with Streptorhynchus.

I am uncertain, but have some reason to believe that it is. In such case this character indicates a most important difference in structure and habit from Schuchertella, whether it can be detected in every specimen or not, since such cementation, as above described, can only coexist with atrophy of the pedicle. At least such an inference seems at present to be unavoidable.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2962?).

STREPTORHYNCHUS PERATTENUATUM n. sp.

Pl. XXIX, figs. 3 to 3c.

Shell very small. The shape of the ventral valve is subelliptical, contracting at the hinge line, and having the longitudinal dimension greater than the transverse. The convexity is rather low and the beak small and depressed. The ventral valve is very high and conical. The aperture is subelliptical and the hinge line shorter than the width in front. The area is rather sharply defined from the sides. The pseudodeltidium is broad, not much elevated, and it occupies fully one-half the width of the area, which is strongly inclined backward. This valve is attached not by a pedicle issuing from between the two valves, but by cementation of the apex of the ventral shell. The growth is irregular. The surface seems to be obscured on the ventral valve, which is nearly smooth. The dorsal valve, however, shows strongly defined liræ, which are somewhat thick and rounded and separated by deep striæ of about the same width as themselves. The liræ are rather fine, coming about five or six in the space of 1 mm. Crenulations are obscure or absent.

This description is based on the typical specimen, and enough material has not been examined to gage the extreme of variation from it. The few additional specimens which have thus far come to hand, however, do not show any notable departures.

I am not sure that this species is more than a much-dwarfed variety of Streptorhynchus gregarium. It comes from a different area, and besides being much smaller has a somewhat different surface ornamentation. The liration is a little finer and the liræ themselves thickened at the expense of the intervening striæ, so that instead of being thin and separated by relatively wide interspaces the liræ and striæ are about equal. Such an appearance, however, is sometimes due to crowding, especially in old shells, and to a certain extent may be produced by silicification.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2962).

STREPTORHYNCHUS? sp. a.

Pl. XXX, fig. 7.

This type is represented by but a single specimen, and as it is a dorsal valve its generic position is of course considerably a matter of doubt.

In size it is small, having a length of only about 5 mm. The shape ascribed to it must depend somewhat on the restoration made, as a guide to which either varices of growth or concentric striæ are almost entirely lacking. The right-hand side, as seen in the figure, appears to be nearly the natural outline, and restored thus the shape would be strongly contracted at the hinge and the width somewhat less than the length. The growth is irregular. The sculpture consists of slender elevated liræ, which are irregular and more or less alternating. Crenulating concentric liræ are faint or absent. The cardinal process is small and the socket plates not prolonged.

The outline, contracting at the hinge, and the internal characters are suggestive of Streptorhynchus and of the species recognized in this report Streptorhynchus? sp. a most closely resembles S. pygmæum, with which it is associated. It does not seem advisable to place it with that species, however, because of several differences. The liræ, while similar in a general way, are thinner and higher, and crenulations and varices of growth appear to be absent. The cardinal process is smaller. The test instead of being massive and mature is thin and possibly immature.

The liration is suggestive of Orthotetes? sp. a, to which an associated specimen has been referred, but the liræ are more irregular and the shape is different. The shape, the sculpture, and the internal structure combined are not found in any other Guadalupian species. Among the specimens from station 2969 identified as Derbya crenulata some approach closely to the form in hand. Most of these have a more or less similar outline, and one has the crenulations very faint or absent, so that it has a pretty close resemblance to the present specimen. Perhaps the example from station 2969 should be withdrawn from Derbya crenulata and placed here, but it does not seem justifiable to assign the present specimen to D. crenulata on its proper characters,

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Genus DERBYA Waagen (emend.).

Waagen proposed the term Derbya in 1884, distinguishing two sections which possess in extreme cases rather striking differences of structure. These are the septati, to which the genotype D. regularis Waagen belongs, and the camerati. Though Waagen uses the term Orthotetes for quite a different group, I find that it was originally employed by Fischer de Waldheim for the same shells for which Waagen proposed Derbya. Fischer even included both types of structure in Orthotetes, though the type species, O. radiatus, belongs to the camerate division. I had been in some doubt as to the advisability of including these two types of structure under a single generic term, and finding two names already in the nomenclature have availed myself of this fact to retain both, the name in each case being restricted to the division to which its genotype belongs. Should a different conclusion be adopted, Orthotetes must clearly supersede Derbya. For this reason Derbya is here used in a restricted sense, as applying only to the septate division of the original generic content, while Orthotetes is employed for the camerati alone. I am, however in some doubt as to the wisdom of separating the septati and the camerati under distinctive names, and have some misgivings lest I may have ventured too far in trusting to the accuracy of Fischer de Waldheim's rather explicit description of the structure of Orthotetes radiatus. The boldest course would perhaps have been the most prudent one, to have eliminated the term Derbya entirely by placing it as a direct synonym of Orthotetes. The course adopted aims to conserve the present terminology as far as possible with adherence to fact.

To Derbya belong the following species described in this report: D. nasuta, Derbya sp. a, and Derbya sp. b. The position of D.? crenulata is doubtful, Several of these types have characters which render them rather striking. D. nastata is remarkable for the height of the septum and the great size of the cardinal process, and D.? crenulata is conspicuous among the forms treated here for the coarseness of its surface ornamentation. A natural grouping of these species is not obvious. No one would think of putting D.? crenulata and D. nasuta into the same group, but Derbya sp. a and Derbya sp. b are not so easily to be disposed of. In fact, Derbya sp. a is itself probably too composite to be at present handled satisfactorily.

The representation of this genus, especially in its most distinguished types, is chiefly outside of the Guadalupe Mountains, in beds supposed to be equivalent to the Delaware Mountain formation. A few examples have been found in the Capitan formation, but there its representation is distinctly subordinate. Orthotetes, which is not known at the lower horizon, is much more abundant.

The Orthotetes of the Guadalupian fauna are in a general way distinguished from the Derbyas by the character and fineness of the liration, Many of the Derbyas do not differ materially in these respects from the common Mississippian and Pennsylvanian species of the Mississippi Valley. Some, however, have fine liræ, and no contrast with the Orthotetes can be drawn which is persistent and absolute.

DERBYA NASUTA n. sp.

Pl. XXVI, figs. 6 to 6c.

I have taken for the type of this species a rather imperfect specimen showing portions of both valves in conjunction, together with some interesting structural features. From this the following description can be framed:

Shell large. Ventral valve semiconical, inclined backward so that the cardinal area makes a rather strongly obtuse angle with the plane of the edge. Area high, flat transversely, with a strong concave curvature longitudinally; width, 50 mm., height, 25 mm. Width of the delthyrium at its base 11 mm. The pseudodeltidium is divided into three distinct portions, the median of which is strongly convex, the lateral portions concave and slightly depressed below the rest of the area. Its lower portion is crossed by imbricating transverse lamellæ. The lateral portions of the area are again subdivided by two diagonal lines situated somewhat nearer the outer edges of the area than the edges of the delthyrium.

The dorsal valve is rather strongly convex and the umbo considerably inflated. The width at the hinge line is 50 mm., but the outline expands rapidly below, and the greatest diameter could not have been much less than 90 rum., while the length must have been not far from 75 mm. Both valves show many irregularities and distortions, due to unequal growth.

The interior of the ventral valve is provided with a long, high median septum, which apparently is not in contact with the pseudodeltidium, except possibly near the apex. The hinge plate and cardinal process of the dorsal valve are extraordinarily large and massive. Their general character is shown in the accompanying illustration (Pl. XXVI, fig. 6c). The most marked peculiarity consists in the strongly diverging and much produced apophyses, which receive between them the septum of the ventral shell. Their general direction is nearly parallel to the plane of the shell edge.

The surface is crossed by moderately thin and high radiating liræ, increased by implantation and having a more or less irregularly alternating arrangement, because of the imperfectly developed intermediate liræ. The spaces between them are about equal to the liræ themselves. The liræ number 8 to 11, usually 10 or 11, in 5 mm., according to the number of young and intermediate ones present.

Horizon and locality.—Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764).

DERBYA? CRENULATA n. sp.

Pl. XXVI, figs. 5 to 5d.

Shell of medium size, subsemicircular. Ventral valve rather low. Growth regular. Surface from apex to front and sides nearly plane. Area directed at about right angles to the plane of the edge. Pseudodeltidium rather strongly convex, narrow, not well defined. Width of area, 30 mm.; height, 5 mm. Width of pseudodeltidium at its base, 2.5 mm. Length of valve from apex to anterior margin, 19.5 mm.

Dorsal valve moderately convex. Beak small, strongly incurved. Umbo rather inflated. Length from beak to front, 18.5 mm.

Surface marked by rather coarse, thin, strongly elevated liræ separated by rather narrow, deep grooves. The liræ come about six or seven in the space of 5 mm., and are crossed by strong crenulations. New liræ are introduced by intercalation, and until they attain full size are somewhat alternating. At intervals also certain of the liræ will sometimes be larger and more prominent, with three to five apparently full-grown but smaller ones between them.

This, the typical specimen, shows strong old-age characters in the superimposed marginal lamellæ. In its youthful stages the hinge extremities were not projecting; at maturity they were very materially prolonged at one side, while the senile stages returned to a form having a quadrate cardinal margin. The interior of the shell is unknown, so that it is impossible to assign it to either the septate or camerate groups, or even with certainty to the genus Derbya. It differs front the common D. crassa of the Mississippi Valley upper Carboniferous in showing a tendency to extension of the cardinal extremities, in having a coarser and stronger liræ and crenulations, and in possessing a somewhat higher area and proportionately narrow pseudodeltidium.

In addition to the type we have specimens apparently belonging to this species from two other localities. Aside from the fact that they are very much smaller, it seems rather probable that they belong to Derbya crenulata. The liration is about the same, though the individual liræ are perhaps a little thinner, and crenulations are very strong. These are all dorsal valves, and do not, therefore, furnish any collateral evidence as to whether Derbya crenulata has a septate or camerate structure.

Horizon and locality.—Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764). Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2969 and 3500).

DERBYA sp. a.

Pl. XI, fig. 1; Pl. X, fig. 11, 11a.

Under this heading is included an assemblage of forms which with complete material might be found to represent more than a single species, but which at present it did not seem advisable to subdivide. In size these specimens present various degrees, from the large dorsal valve, represented on Pl. XI, fig. 1, to diminutive examples having a width of only 6 or 7 mm. The latter are of course presumed to be but immature representatives, whose adult growth would have brought them to the dimensions found in the others.

Generally speaking, the dorsal valves have a low convexity, with the hinge line equal to the width below. Ventral valves have a similar outline. The area is not as a rule very high. While in some specimens it is about perpendicular to the margin of the valve, in others it is inclined backward, and in still others forward. Growth appears to have been more or less irregular, though some specimens are symmetrical. The pseudodeltidium is broad and not strongly elevated.

The surface ornamentation consists of rather fine liræ, of which about 18 to 20 occur in a space of 10 mm. The liræ are naturally thin and separated by relatively wide striæ, but by exfoliation they are liable to appear rounded and somewhat obscure. Crenulating concentric liræ are faint, if present.

On the inside the ventral valve has a strong median septum, which appears to be normally distinct from the dental plates and which in no specimen has been observed to be connected with them so as to form a well-defined chamber. In the dorsal valve the cardinal process is relatively small and the socket plates somewhat stout and prolonged.

The representation of this species is scattering, and in no instance have a ventral and a dorsal valve been found in conjunction.

The large dorsal represented by my figure has a length probably exceeding 57 mm., while the greatest width must have been 80 mm. or more. The shape is transverse, subelliptical, the width at the hinge being as great as at any point below. Convexity is only moderate.

The surface ornamentation is largely obscured by exfoliation, but the liræ appear to be fine and not very distinct. A small area in a better state of preservation shows them to be moderately thin and high, with interspaces of about the same width. There are about 17 to 19 in a space of 10 mm.

The cardinal process is bilobed, moderately extended, and directed somewhat downward when the shell is viewed from the convex side.

The size of this specimen and the surface ornamentation are suggestive of the form here described as Derbya nasuta, but it is of more regular growth, shorter at the hinge line, and with a much shorter and smaller cardinal process. It also recalls Derbya robusta Hall, but it is larger, less strongly convex, and wider at the hinge.

Associated with this was found another smaller fragmentary dorsal valve, and a small ventral valve which is represented by figs. 11 and 11a of Pl. X. Here the size is rather small and the growth unequal and unsymmetrical. The area is generally flat, moderately high, and backward inclined. Its width is about 12 mm., considerably less than that of the shell in front, and the height is 4.5 mm. The pseudodeltidium is elevated, somewhat flattened, 2 mm. wide at its base. The length of the shell from apex to anterior margin is 19 mm., from cardinal line to anterior margin slightly less. Width 21 mm. The septum has a length of only 8 mm.

The surface is marked by more or less angular and spaced, rather strongly elevated radiating liræ, of which 11 or 12 occupy the space of 5 mm. The liration, though perhaps imperceptibly finer, gives the exterior of the shell an appearance so exactly like that of the large dorsal valve with which it was found associated that it seems not wise to separate them, yet the contraction of the outline at the hinge causes me to doubt the propriety of placing it with the other. This shell differs from Orthotetes guadalupensis and its allies in the coarser liration and also its lower area, broader pseudodeltidium, etc. Other ventral valves placed here have relatively wider hinge lines and in some cases more regular growth.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906; "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2962 and 3501).

DERBYA sp. b.

Pl. X, fig. 12.

This species is primarily represented by a single ventral valve from the Capitan limestone. The shape is elevated, conical. The area is sharply defined from the anterolateral curvature and generally nearly flat. Its width is 16 mm. and its height about 12 mm. The pseudodeltidium is wide (about 4.5 mm. at its base) and moderately convex.

Because the anterior part of the shell is imperfect the shape of the aperture can not be definitely determined, but it appears to be semicircular and strongly transverse. Restored as nearly as possible, the plane of the aperture is perpendicular to that of the area, and its length is about 10 mm., its width being of course the same as the width of the area, or 16 mm. The growth is irregular.

The surface is marked by varices of growth and by rather coarse concentric striæ. The liræ are distinct and sharp, but low, the intervening striæ being shallow, rounded, and traversed by the concentric striæ. There are about 9 liræ in 3 mm.

The internal structure of this shell indicates that it should be considered one of the septate Derbyas. The septum is high and independent of the low dental lamellæ. It has some appearance of being composed of two plates, which are in contact and cemented one to the other for most of the way, but separate a little near the anterior wall. One might possibly be justified in regarding this shell as an abnormal Orthothetina, though that view is not adopted here.

This form recalls Orthotetes guadalupensis, but the liræ are too angular and the pseudodeltidium too wide. In the latter respect it resembles Orthotetes distortus, but the liræ are much too coarse. Indeed, it does not seem to possess the camerate structure which is found in this group of forms. It seems to be clearly distinct from Derbya sp. a and from Derbya crenulata, nor do I feel that it could safely be united with Derbya nasuta.

I have also referred to this species a small silicified shell from station 2969. It is fragmentary, but in every particular, so far as its characters are shown, it agrees with the original specimen from the Capitan limestone. Perhaps one difference may be found in the fact that the well-developed septum consists of a single homogeneous plate. This shell possesses the feature—rare for a Derbya—of showing a large and unmistakable scar left by apical cementation. As the apex of the original specimen is broken, however, this may be a character of identity rather than a character of difference between them.

This specimen is associated at station 2969 with certain small shells of somewhat similar dimensions, which have been placed with Derbya crenulata. The circumstance that they occur associated is suggestive that they may represent the same species, but the ventral valve here under consideration has finer, rather obscure, and apparently uncrenulated liræ, from which I am led to infer that it is not conspecific. If it does belong with the dorsals, the latter can not be referred to Derbya crenulata, where I have placed them, nor probably to the present species.

Horizon and locality.—Top of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2966). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Genus ORTHOTETES Fischer de Waldheim.

It seems almost certain, after careful examination into the history of the names, that Fischer de Waldheim's term Orthotetes, which Waagen employs with quite a different force, was originally applied to shells having the characters that distinguish Waagen's genus Derbya. The history of this genus, so far as it is known to me, is as follows:

The term Orthotetes was first introduced in the year 1829,a The following is a complete reproduction of the original description of this genus:

---

aBull. Soc. imp. des naturatistes de Moscou, vol. 1, 1829, p. 375.

---

TRAVAUX.

Le directeur a disserté sur quelques fossiles du gouvernement de Moscou.

?Gryphaea biceps. an Inocérame Brongn. ?
Strigocephali spec.
Strophomenæ spec.

Il a rendu surtout attentif sur une coquille bivalve que Mr. Évans a trouvée à Pakhrino at qu'il croit d'un genre nouveau, qu'il a nommé: Orthotetes, du grêc ; droiture, parce que la charnière présente une impression transversale, droite et linéaire. On n'en connait malheureusement qu'une valve.

Voici las caractères génériques qu'il lui assigne:

Coquille libre, subrégulière, plate, subéquivalve, subéquilatérale. Charnière droite et transversale. La valve operculaire, offre une impression articulaire droite interrompue an milieu par un enfoncement profond, qui est couvert par un prolongement de la charnière comme par un toit qui avance même au delà du plan de la valve et parait s'adapter à la valve supérieure.

Cette apophyse on [ou] ce prolongement avance en dedans de la coquille en une arête droite et canaliculée.

L'enfoncement pour l'attache du muscle est très grand, circulaire et rayenné. Le dos de la charni&egrav;re et [est] applati, lisse et comme scié et poli.

Cette coquille offre quelque ressemblance avec les genres Placuna, Pedum, etc. Il y a même un canal creusé, dans le dos de la charnière, ce qui peut montret [montrer] quelque analogie avec la dernière. Elle est au reste presque aussi mince qu'une Anomie, mais très regulièrement aplatie et rayennée.

This description was unaccompanied by figures, and no species were cited under the genus thus newly established.

The year following, Fischer de Waldheim published another reference to the genus a in the form of three figures, which are reproduced in Pl. IV (figs. 1, 1a), of the present work. The only text relating to these figures which can be connected with the publication of the previous year is the following, the description of Pl. 20, fig. 4.

---

aOryctographie du gouvernement de Moscou, 1830, pl. 20, figs. 4a and 4b.

---

4. Orthotetes, nouveau genre de coquilles bivalves. v. Bulletin de la Soc. imp. des naturalistes de Moscou, 1829, p. 375.

The next reference to the genus was in a later edition of the work last cited, published in 1837.b

---

bOryctographie du gouvernement de Moscou, 1830-1837, pl. 20, figs. 4a, 4b, and 4c.

---

I have seen it claimed in a book catalogue that the current statement that there are two editions of this work is erroneous, and that it appeared in parts or fascicles from time to time. The actual truth, as is so often the case, probably occupies halfway ground. Two copies of this work are found in the library of the United States Geological Survey. One bears as the date of imprint the year 1830. It contains an advertisement by Fischer de Waldheim dated March 23, 1830, plates to the number of 66 (maps and cuts, Pls. A—G; fossils, Pls. 1-44; natural history, Pls. 1-15), and brief explanations relating to them, but no text. The imprint of the other volume is dated thus: "1830 = 1837." It contains a similar advertisement, somewhat changed, but over the same date, a table of contents (pp. xi-xvii), descriptive text (pp. 1-189), explanations of plates (pp. 191-195), an index (pp. 196-202), 7 maps and cuts (Pls. A—G), and 51 plates of fossils (Pls. 1-51), those on natural history having been omitted, presumably to make the contents properly fit the title. There have been not only additions, however, but changes. The advertisement has been set up afresh; the typography is different, the pagination is different, and the matter, while repeated word for word for the most part, has at least one important textual change. The description of the plates has also been changed, being now arranged on the page in two columns instead of one, and some of the matter itself has been altered. Two of the plates even have been changed, by the insertion of additional figures, one instance of this being important to the subject in hand. The conclusions which I draw from these facts is that the bulk of the plates, with explanations but without text, were published in 1830; that additional plates and parts or all of the text were issued during the succeeding seven years; and that in 1837 was published a final revised edition of the whole.

In this volume figs. 4a and 4b are reproduced from the edition of 1830, and an additional figure (4c) is inserted, which appears as fig. 1b of Pl. IV in the present work. The explanation of the plate reads:

PL. XX.

1. 2. 3. Inoceramus concentricus, BRONGN., p. 177.

4. Orthotetes, n. g., p. 133. valve inférieure de Pintérieur.

(a) charnière applatie vue en dessus,
(b) canal dorsal de la même valve,
(c) valve superiéure.

5. 6. Strophomena Pecten, p. 145.
Orthis Pecten, DALMANN.

7. 8. Strigocephalus Defrancii, p. 145.
Sphaera corrugata, Sow., Pl. IV, p. 335.

The complete text relating to this genus, which, it may be remarked, Fischer de Waldheim includes along with Chonetes among the pelecypods, is as follows:

ORTHOTETES.
Pl. XX, f, 4, a, B, c.

Coquille bivalve libre, subéquivalve inéquilatérale; charnière droite transverse, lisse.

La valve operculaire présente une impression droite, qui est interrompue par un avancement de la charni&egrae;re, en forme de tablette carrée, creuse intérieurement, prolongé intérieurement par une apophyse médiane et droite.

Deux impressions musculaires presque circulaires à côté de l'apophyse médiane.

Orthotetes, FISCHER. Bulletin de la Soc., 1829, tome I, p. 375.

Le mom est déduit du grec, , droiture, faisant allusion à la forme droite de la charnière.

Je ne connais rien de semblable à cette forme de la charnière, difficile à rendre claire par la description. On peut dire que la forme totale peut rappeler celle d'un peigme sans ailes, qu'elle est rayennée comme une Strophomène, qu'elle a une impression dorsale sur la valve operculaire comme Pedum, et qu'elle est presqu'aussi mince comme une Anomie.

La forme de la charnière avec ses apophyses intérieures paraît rapprocher cette coquille des Brachiopodes.

Cette coquille so distingue de toutes les autres bivalves, par la charnière droite, mince et édentée dont le sillon en dessous reçoit le bord mince de la valve opposée. Le prolongement carré du milieu dépasse de beaucoup la charnière et cohère dans l'intérieur de la coquille avec une longue apophyse droite, qui partage le creux musculaire. La valve opposée ou dorsale n'a qu'un endroit bombé près de la charnière, elle est au reste plate dans tout son pourtour. M. EVANS a trouvé cette coquille le premier dans le calcaire de Kalouga; je l'ai retrouvé dans celui de Podolsk.

The fourth important publication relating to the genus appeared in 1850.a This is quoted in full as follows, and the accompanying figures are reproduced in Pl. IV.

---

aBull. Soc. imp. des. naturalistes Moscou, vol. 23, pt. 1, 1850, pp. 491-494.

---

ORTHOTETES.
Genre de la Famille des Brachiopodes restitué par G. Fischer de Waldheim, avec Pl. X.

Lorsque j'ai produit le genre Orthotètes en 1829 (Bulletin de la Soc., I, p. 375), mes connaissances n'étaient pas assez claires pour en fonder les caractères génériques. J'ai reproduit ce genre dans l'oryctographie de Moscou, accompagné d'une figure (1837, 133, Pl. XX, f. 4), mais j'avoue franchement que je n'y ai rien ajouté à la vraie connaissance du genre.

Maintenant, grâce aux recherches géologiques de M. Vosinsky qui a trouvé dans le district de Serpoukhof de notre gouvernement une seconde espèce, je suis à même de compléter les caract&egrav3e;res de ce genre.

M. d'Orbigny dans le second volume de l'ouvrage sur la Russie de Murchison a placé cette coquille sous le genre Orthis. Il faut convenir que les deux genres présentent une grande affinité. Mais l'Orthotètes so distingue déjà extérieurement de l'Orthis, par la grandeur, par la compression totale de ses valves, et par la forme arrondie depuis la charnière jusqu'au bord opposé, par les stries élevées rayonnantes et surtout ensuite par sa charnière.

Caractères gènériques du genre Orthot`tes.

Testa bivalvis, subæquivalvis plana valde compressa cardine dentato, dentibus binis latis acuminatis, f. 1.

Area recta, lævis, plus minusve lata, deltidio oblongo, apice obtuso, f. 2.

Valve dorsalis intus appendicibus donata.

La forme large et déprimée de la charnière la distingue au premier coup d'œil de l'Orthis. L'une des espèces ne montre qu'une petite élévation ou bosse sous la charnière, qui dépend d'une espèce de charpente intérieure et quadrangulaire. Cette petite caisse est collée à la valve, ainsi que l'appendice longitudinale médiane. Cette appendice est accompagnée de deux autres obliques, qui forment avec la médiane une espèce de croix. Dans l'autre espèce l'appendice médiane est simple, sans les branches latérales.

Voici la description de ces deux espèces.

I. Orthotetes radiata.
Pl. X, f. 3.

Testa plana striata, striis elevatis radiatis.

C'est cette espèce qui montre a l'int&egraave;rieur ces appendices dont ii est question plus haut. Elle est très comprimée et ne montre que sous la charnière une petite bosse qui couvre cette charpente carrée intérieure. Deux sillons circulaires se trouvent pres du bord et le suivent dans son contour.

Orthis arachnoidea, d'Orbigny-Murchison. Géologie de la Russie, II, p. 196-197. Pl. 10, f. 18, a, b. Pl. II, f. 1, a, b.

La description en est faite de main de maitre et il faut la comparer. C'est la même coquille. Mais l'Orthis ou Strophomena Pecten de l' Oryctographie (Pl. XX, f. 5, 6) en est différente; elle s'en distingue par sa charnière dilatée et par son épaisseur. Elle se trouve dans les couches supérieures d'un calcaire ferragineux de Dorogomiloff près de Moscou.

L'orthotètes rayonné existe isolement dans les couches inférieures du calcaire carbonifère de Podolsk du Gouvernement de Moscou et près de Médine du Gouv. de Kalouga.

2. Orthotetes socialis.
Pl. X, f. 4.

Testa plana valde compressa, tenuis fere papyracea, striata: striis tenuissimis radiatis.

Cette coquille est plus petite quo la première. La hauteur depuis la charnière jusqu'au bord est d'un pouce six linges; la largeur de l'Area comprend un pouce quatre lignes.

M. Vosinsky, membre de notre Société, connu par ses recherches géologiques, a trouvé cette espèce dans les couches infèrieures du calcaire carbonifére dans le district de Serpoukhoff, à Lisenki, propriété de M. le Prince Wiazemsky (Gouvernement de Moscou). Elle est répandue dans un calcaire schisteux qui s'étend à plusieurs verstes et s'y trouve l'une couchée contre l'autre, ou souvent l'une couvrant Pautre. Elle est tellement mince que les deux valves no dépassent pas l'epaisseur d'une ligne.

Les stries sont extrêmement fines et rayonnées. On y observe quelquefois deux sillons parallèles au bord.

These four publications of Fischer de Waldheim include, so far as I am aware, all the authentic data relating to the genus Orthotetes. Lacking the typical specimens, we must try to decide from these four works what the characters of the genus really are, and to this consideration I now proceed. The other publications to this genus, which will be briefly considered later, are important chiefly for historical reasons. It is possible that one of these (Trautschold) actually dealt with the same species which Fischer de Waldheim investigated, but the others offer discussions of Orthotetes apparently based upon species belonging to other genera.

The opinion seems to have been common (Waagen and Trautschold) that Evans was the real author of this genus, and that Fischer de Waldheim acted only as an agent in adopting or introducing the name. The point is not a material one, as in any event there could hardly be any questions as to the technical authorship. It seems to me, however, entirely unnecessary so to interpret the statement in the publication of 1829. I understand that nothing more than the finding of the specimens was ascribed to Evans, and that "croit" and "nommé" do not refer back to "Mr. Evans," but to the subject of the whole sentence, "le directeur," Fischer de Waldheim.

It is evident also, from the derivation cited by Fischer de Waldheim, which should be written , that the only correct orthography of this name is "Orthotetes," and not, as usually seen, "Orthothetes."

It is also evident that Fischer de Waldheim did not at first give an adequate description of the genus according to present ideas; that not only was his description inadequate but that he did not name a type or cite any species from which a type could be selected; and that he even misconceived the structures which his fossils showed. What these characters really were appears with progressive clearness in subsequent publications by the same author, and it is to the consideration of this point that I shall now proceed.

One can safely say that in Fischer's own mind many of the characters of the genus were not clear, and that from this circumstance, from his misconceptions and false analogies, and from the lack of a definite terminology, many of his statements convey either an untrue meaning or no meaning at all. At the same time I feel satisfied that from the first description this genus was founded on a ventral instead of a dorsal valve, as has usually been declared, and that it contained a septum which probably connected with the dental plates to form the typical structure of Waagen's camerate division of "Derbya."

One must bear in mind that in these earlier publications Fischer de Waldheim thought this to be a pelecypod genus. Had he known both valves it is less probable that he would have entertained this idea, and in fact he states in the first description that only one valve is known. His comparison of either valve with Anomia and Placuna, at least as these genera are now understood, shows what a very imperfect and fantastic conception the author must have had, both of Orthotetes and of those types as well. The key to the whole matter seems to me to rest in his comparison with Pedum, a pectinoid genus having a large cardinal area with an elongate median pit for the ligament. There can be no question which valve of a strophomenoid brachiopod presents the closest similarity to this structure. It is true that in "Derbya" and its allies the cardinal structures of the dorsal valve are analogous to those of the ventral. The dorsal area, however is always extremely narrow. It has no proper delthyrium a and chilidium. By the reduction of the area the delthyrium is reduced in size, and it is closed by the development of a large cardinal process, at the base of which, on its outer side, the chilidium occurs as two little ridges more or less connected into a continuous transverse band. The presence of a septum in the dorsal valve is, according to my observation, very rare, and it is usualally obscure. It seems to me that these structures in the dorsal valve are so inconspicuous as to have escaped any but a closer scrutiny than it is probable Fischer de Waldheim ever gave to his specimens, while in the ventral valve they could hardly be overlooked.

---

I can not find that this structure in the dorsal valve has ever received a separate name, and have been forced to employ a term which seems to pertain to the ventral.

---

See how his description answers to this interpretation. The shell is described as "flat," an expression much more appropriate for the ventral than for the dorsal valve. At the same time his remarks here seem to apply to the genus as he supposed it to be in its entirety, both valves combined, for he immediately proceeds to discuss the "lower" valve. So I understand his expression "operculaire" (operculated, or provided with a cover), and in fact he subsequently figures his specimen as a "lower" valve. This, he says, has a straight cardinal area "impression articulaire") interrupted in the middle by a delthyrium ("enforcement profond") covered by a prolongation of the hinge "comme par un toît," which must be a pseudodeltidium. The area again seems to be referred to in the words "dos de la charnière," and "arête droite" is evidently a median septum whose relation to the pseudodeltidium is misconceived and not clearly described. The distinction is perhaps made between the cardinal line ("charnière"), described as "linéaire" and the area, for which the expression impression "articulaire" is used. Had this been a dorsal valve such a description could hardly have been given. The adjective "linéaire" is not applied, as I interpret it to the cardinal area, nor, I believe, can the term "apophyse" be taken to mean a cardinal process (which would immediately settle the status of the valve under discussion), since it projects inward and connects with the septum.

These interpretations are fully borne out by the illustration published in the Oryctographie the year following. Here we are clearly shown from the view of the interior that the shell had a strong median septum, and from the elevation above that it had a high area and pseudodeltidium entirely incompatible with the identification of this as a dorsal valve. The figure leaves much to be desired in representing the relation of the septum to the area and pseudodeltidium, but this ambiguity is quite in harmony with that of the language employed the year previous in describing the same structures. Certain features are clearly out of drawing in the illustration 4a, and the other cut (4b) is almost unintelligible, though it may represent the area, septum, and flabelliform muscular scar. In the edition of 1837, however, this is described as being the "dorsal canal," whatever that may signify. These figures at least make it clear that Fischer de Waldheim had in hand a ventral and not a dorsal valve, and that it belonged to the group for which Waagen proposed the name "Derbya," having within a single strong median plate or septum. Here again it will be noted that the figured specimen is not identified specifically, so that the genus is still without a quotable type.

The same is true of the publication of 1837, in which the genus is defined in about the same words as seven years previously. It would, indeed, appear, as Fischer de Waldheim remarks, that the structures of the hinge are "difficile à rendre claire par la description," but in his words "impression droite" I recognize the cardinal area interrupted by "une avancement de la charnière" (pseudodeltidium covering the delthyrium). This ("en forme de tablette carrée, creuse intérieurement"), "prolonged within by a straight median apophysis" (which is a median septum), describes, if I am not much mistaken, the structure of Waagen's camerate group of Derbya. It will be remarked that his comparisons to the pelecypod genera Pedum, Placuna, and Anomia are less insisted upon. Placuna is omitted altogether. The resemblance to Pedum would seem to rest on its dorsal muscular impression and to Anomia on its delicate shell. Resemblance to the structures of brachiopods is noted. The figures, two of which were reproduced from the first edition, yield less additional knowledge than the text, where some changes were in fact made. The third figure is a cut of what is called the "upper valve," which is described as being flat except for an inflated region near the hinge. This might be said either of the dorsal valve or of a not quite regularly developed ventral, but the figure looks like the latter. As at this time Fischer de Waldheim regarded the genus as belonging to the Pelecypoda, and therefore being nearly equivalve, it would not be unnatural for him to assume two somewhat differently shaped ventral valves to be one the upper and the other the lower shell of his supposititious genus, especially since one was known only from the outside and the other from the inside.

In Fischer de Waldheim's final discussion of the genus it will be observed that at last two species are described and definitely referred to Orthotetes. The usual procedure would be to use for a type that called Orthotetes radiatus, no type being named and that species standing first under the description. Of the four figures accompanying this publication only two are cited in connection with any specific name, the two others being mentioned in connection with the generic description. Thus fig. 3 represents Orthotetes radiatus and fig. 4 O. socialis, while fig. 1 is supposed to show the hinge teeth and fig. 2 the cardinal area. Anyone comparing the figure of O. radiatus (fig. 3) with the original figures of the Oryctographie, for which an opportunity is here presented, must become convinced that it is merely a more faithful representation of the same specimen, while fig. 2 is evidently but an improvement of another of the original figures, that showing the cardinal area of the same specimen. It beyond question accompanies fig. 3 in belonging to Orthotetes radiatus. It seems to me clear, therefore, that O. radiatus, which would naturally be used for the genotype of Orthotetes, is based on the same specimens which furnished the original description and figures of the genus before any species belonging to it had been given names. That O. radiatus is really the species originally used by Fischer de Waldheim in describing Orthotetes is shown by the remark that by means of a second species, found in the district of Serpoukhoff, he was able to supplement the characters of the genus. The second species thus referred to is O. socialis, found by M. Vosinsky in the "Lower Carboniferous" limestone of the district of Serpoukhoff, while O. radiatus is the original species, as the figures indicate.

For the species which must be taken as the type of Orthotetes Fischer de Waldheim uses the name O. radiatus, but it is somewhat doubtful whether this name can be so applied. In the first edition of the Oryctographie, on the same plate as the Orthotetes figures, this author represents a shell from Dorogomilov, near Moscow, under the name Strophomena radiata. In the 1837 edition these figures are reproduced, but because he believed his shell to be the same as Dalmann's Orthispecten he drops his own name, Strophomena radiata, and cites the species as Strophomena pecten, but later he says, as quoted above, that Strophomena pecten from Dorogomilov is distinct from the type of Orthotetes, and that it occurs in higher beds. It is evident, therefore, that this form from the Russian "Coal Measures" is generically distinct from the Ordovician species Orthispecten Dalmann, and that for its specific name we must fall back on radiata, used for it in the first edition of the Oryctographie. Unfortunately, as it has already been seen, Fischer de Waldheim revives this specific name for the type species of Orthotetes, which he says is distinct from Strophomena? radiata. The appearance of his figures is exactly that of an Orthotetes or a Derbya, but might be equally well an Orthothetina or a Schuchertella, etc. Of course if Strophomena radiata Fischer de Waldheim 1830 is found to belong in the same genus as Orthotetes radiatus Fischer de Waldheim 1850, another name will have to be proposed for the type of Orthotetes, unless an available one can be found in the synonymy. As to this I am not sufficiently familiar with the synonymy of Russian forms to reach a conclusion, and doubtless recurrence would have to be made to original specimens, which for me is of course impossible. I therefore use the name O. radiatus for the type species of Orthotetes, conditionally on the contingencies above defined.

In the publication under consideration (that of 1850) figs. 2 and 3 belong to Orthotetes radiatus; fig. 4 is definitely ascribed to Orthotetes socialis, with which fig. 1 also should probably be associated, a point to which I shall recur. On the other hand, while the twin figures which appear in the Oryctographie merely as fig. 4a are represented in the later publication as figs. 2 and 3, fig. 4c, which purports to represent the "upper valve," has been suppressed (possibly because it is really a ventral), which is also the fate of 4b, described as the "dorsal canal" of the "lower valve," unless it appears as fig. 1, which I have otherwise disposed of as belonging to O. socialis.

Let us now consider what are the characters of Orthotetes as based on the latest and most intelligent authentic account of the genus. Two types of structure are here discriminated by Fischer de Waldheim. One species is described as having the median septum ("appendice longitudinale mediane") connected with two other plates oblique to it, forming a sort of internal quadrangular chamber ("caisse").a In the other species the median septum is simple, without lateral branches. This language, while not technical, is quite plain and clearly describes in the first instance the type of structure on which Waagen based the camerate division of his genus Derbya, while the other describes his septati. Since under O. radiatus, described immediately thereafter, Fischer de Waldheim says "It is this species which shows on the inside the lamellæ ['ces appendices,' by which term are designated both the septum and dental lamellæ in the remarks quoted] referred to above," it seems clear that it is O. radiatus which possesses the internal chamber, a structure which is imperfectly represented also by his figures of that species, while O. socialis has the structure of the septati. It is clear that the interior of the latter species was known to Fischer de Waldheim, because one of his first remarks is that it permits him to complete his generic diagnosis, and because he proceeds to give characters belonging to the internal structure. It would be natural, therefore, that he should illustrate the interior of O. socialis, as well as the exterior, which is shown in fig. 4, and it seems highly probable that fig. 1, which is not ascribed to either of the species, and which appears to represent a shell belonging to the septati, really serves this purpose, and goes with fig. 4 as illustrating O. socialis.

---

aIt is this structure, here for the first time described and figured in a fairly intelligible manner, which is referred to as an "apophyse" or "prolongement" of the pseudodeltidium (toît, vide context), prolonged within as a median septum (arête droite et canaliculée), and which is somewhat correspondingly represented in the 1830 edition of the Oryctographie. The camera combined with the pseudodeltidium is evidently meant by the words of the 1839 edition, "un avancement de la charnière, en forme de tablette carrée, and in the work under consideration is first mentioned as "une petite élévation ou bosse" beneath the hinge, being described more fully immediately thereafter.

---

To consider now the work of certain other authors as it bears on the history or the proper understanding of this genus, the remarks of De Verneuil in the "Geology of Russia" (1845) first call for comment. This author places Orthotetes among the Brachiopoda, so that when this step was taken by Fischer de Waldheim five years later he was merely accepting the views of the distinguished Frenchman, but De Verueuil confuses the group with Orthis, of which genus he makes Orthotetes a synonym. Indeed, he places both Orthotetes Fischer de Waldheim and Strophomena pecten in the synonymy of Orthis arachnoidea Phillips. His description of the interior seems to be based on Fischer's figures as much as on actual specimens, and apparently applies to the "septate" type of structure. To his description, however, much weight can not be attached so far as it stands against Fischer de Waldheim's, unless it is shown to have been derived from more authentic or more accurately identified material than would appear to be the case. Fischer, as we have seen, while acknowledging the correctness of De Verneuil's description as a whole, yet maintained the independence from Orthis of his genus Orthotetes, as was quite justified, and described as new the two species Orthotetes radiatus and O. socialis.

We find in 1873a De Koninck accepting Orthotetes as a valid genus distinct from Orthis, but his knowledge of the great diversity of forms which he assigns to it under a single specific designation seems to be restricted to the external configuration, not having extended to the structure of the dental and septal plates. He therefore uses Orthotetes to replace Streptorhynchus and overlooks all the internal differences which have since been used for generic discrimination among the unplicated Orthotetinæ. In recognizing Orthotetes as distinct from Orthis, De Koninck makes no advance, however, since Fischer maintained the same distinction in his work last quoted, while English paleontologists had for many years separated the shells under the name of Streptorhynchus, which, as we have just seen, De Koninck regards as the same as Orthotetes.

---

aMonographie des fossils Carbonifères de Bleiberg, en Carinthie: Recherches sur les animaux fossiles, pt. 2, 1873, p. 42.

---

In 1876, or three years after De Koninck's description, Trautschold,b in describing the Carboniferous fauna of Mjatschkowa, had occasion to discuss some of the species on which Fischer de Waldheim worked. He places Orthotetes radiatus in the synonymy of Orthis crenistria and represents that species by a figure which one would say clearly showed the structure of the camerate group of "Derbya." Schellwien c refigures Trautschold's specimen as Derbya sp. The somewhat better figure given by Schellwein is not altogether unambiguous, but I have little doubt that it represents one of the septati. Mjatschkowa is the locality from which this specimen was obtained. Orthotetes radiatus was first cited from Pakhrino, a small hamlet south of Moscow. In the publication of 1850 Fischer de Waldheim mentions as the locality Podolsk, a larger town near Pakhrino and somewhat west of it. Pakhrino is situated about midway between Podolsk and Mjatschkowa. It is not improbable therefore that Trautschold's specimens actually belong to Orthotetes radiatus.

---

bDie Kalkbrüche von Mjatschkowa, Moscow 1876, p. 63.
c Neues Jahrbuch, Jahrg, 1900, vol. 1, 1900, pl. 1, fig. 9.

---

A great step in advance was marked by Waagen's monograph on the Salt Range fossils,a in which he employs septal structures for the classification of this group of shells. As we have seen, even after the later description of Orthotetes, which, indeed, if not quite adequate was sufficient to distinguish it from Orthis, Fischer de Waldheim's genus did not find general acceptance and apparently was for the most part quite overlooked or where recognized was confused with Orthis, as by De Verueuil and Trautschold, or made to include most of the Orthotetinæ, as by De Koninck.

---

aMem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, pp. 575 et seq.

---

Waagen's remarks on this genus are as follows:

Genus: Orthothetes. Fischer v. Waldheim.

Though this name has been quoted as applied by Evans to certain forms already in the year 1829, yet the genus can not be considered as fairly established before the year 1830 when in the first edition of the "Oryctophie" the interior of a dorsal valve was distinctly figured and the genus definitely transferred to the Brachiopoda by Fischer v. Waldheim. In the edition of 1830 only the interior of the dorsal valve was figured, whilst in the edition of 1837 an external view of a ventral valve is added.

In both cases there can not remain the slightest doubt that the name was applied to a shell very nearly related to Streptorhynchus crenistria, Phill., and which chiefly in the internal characters of the dorsal valve is generically identical with Phillips's species. The name must thus be restricted to those forms and the genus may be characterized in the following manner:

The external shape of the shells belonging to this genus is in no way characteristic and it is only by the internal characters that the genus can be recognized.

In the dorsal valve a cardinal process of moderate dimensions exists, which is generally bifid and comparatively broad. Laterally it is joined to the walls of the dental sockets. These latter are not very large and not supported by shelly lamellæ, so that the diverging septa which are characteristic of the preceding genera are absent in the present one. Instead of these, however, there seems to be not rarely a median dorsal septum developed which is immediately joined to and takes its origin at the cardinal process. This median dorsal septum appears on Fischer v. Waldheim's original figure as well as on Pl. XXVII, fig. 6, of Davidson's Carboniferous monograph.

In the ventral valve as far as my experience goes every kind of septum is absent.

The muscular impressions of both valves have been excellently described by Mr. Davidson, but they are in their general arrangement very similar to the muscular impressions of other genera.

There can be no doubt that Strept. crenistria, Phill., belongs to this genus even if it should be proved that Fischer v. Waldheim's original specimen did not belong to the same species. Also the species or varieties St. radialis, Phill., and St. arachnoidea, Phill., perhaps, too, the typical St. cylindrica, McCoy, will have to be connected to it. Otherwise not much is known of species of the genus.

There can be no question that the spelling of the name "Orthothetes," which appears to have originated with Waagen is without authority or argument. The implication that the name was first employed by Evans also seems to me unwarranted. It is, furthermore, quite clear that the genus was not recognized by Fischer as belonging to the Brachiopoda in 1830 nor in 1837, but much later, in 1850, after it had been so determined by De Verneuil. Waagen also commits an error which seems to have begun with De Koninck, in saying that Fischer's original figures represent a dorsal valve. A careful comparison of Fischer's texts and figures, as I have attempted to show, must convince an impartial student that the shell was a ventral valve. This misapprehension on the part of Waagen is so fundamental that for once this excellent paleontologist has been led quite astray and his diagnosis based on Streptorhynchus crenistria and its allies is erroneous.

Hall and Clarke, the latest contributors to the history of the genus, unfortunately follow Waagen, without having referred to the original sources. Indeed, these publications are so rare that but few opportunities for making the necessary comparisons exist in this country. These authors give an excellent discussion of the group of forms to which Waagen misapplies the name Orthotetes and under Derbya an equally adequate discussion of the group to which it more properly belongs, but naturally add nothing to the present subject. They do indeed call attention to the fact that in his 1850 publication Fischer de Waldheim figures a species of Orthotetes which shows the structures of "Derbya," but they fail to note that this is the typical species of Orthotetes and that the figures are drawn from the same specimens from which the original description of the genus was derived.

It seems to me that the evidence submitted shows almost beyond dispute that the shell first described by Fischer de Waldheim was a ventral valve; that the same specimen was used later to illustrate his improved diagnosis of Orthotetes and to found his species O. radiatus; that he was consistent in describing a second species of Orthotetes possessing essentially the same structures; that though he took cognizance chiefly of structures calculated to distinguish Orthotetes from the Orthidæ, and failed to lay stress on such as would distinguish it from other groups of the Strophomenidæ, these distinctive characters are more or less clearly described in his text and represented by his figures from first to last; and that while Waagen proposed the name Derbya for the structural type to which Orthotetes really applies, he employs the name Orthotetes for a group having the structures neither as described by Fischer de Waldheim, nor as shown by the type species O. radiatus. Thus Derbya of Waagen becomes, strictly speaking, a synonym of Orthothetes and the group which Waagen discriminated under the name Orthothetes becomes anonymous. For this group I proposed the name Schuchertella, in honor of my one time colleague, taking for the type of the genus the Kinderhook species Orthotetes lens White. As I shall elsewhere (p. 198) have occasion to remark, there seems to be some doubt as to the structure of Orthotetes? crenistria, the species commonly employed as the genotype of Orthotetes, so that Phillips's species could not safely be selected for this office. On the other hand, the interior of Schuchertella lens is accurately known from specimens undeniably belonging to that species. The characters of the group for which the name Schuchertella was proposed have been described not only by Waagen, but more recently by Hall and Clarke,a who also illustrate the type species S. lens.

---

aNat. Hist. New York, Pal., vol. 8, pt. 1, 1892, p. 253.

---

It would appear that as O. radiatus seems to represent the camerati and O. socialis the septati, the genus Orthotetes in its original content is precisely equivalent to Derbya. But the typical species of Derbya is one of the septati b while the typical species of Orthotetes is one of the camerati. Since the difference between the camerate and septate structures when characteristically developed is somewhat striking, I avail myself of this circumstance, by restricting each name to the division which its typical species represents, at the same time to give independent existence to these two groups of shells and to retain both of the current names. The term Orthotetes, therefore, is here employed only for the camerate division of the original genus Derbya, while Derbya is restricted to its typical division, the septati.

---

bWhile Waagen had Derby and his Brazilian species especially in mind, in naming the genus Derbya, the only Indian shells which he discovered belonged to the septate division; and shells of this type were for a long time the only ones which were well known. Waagen did not designate a type species for Derbya, and probably the first one described under that genus, Derbya regularis, should be used. This species has been employed in this sense by Hall and Clarke, and as it belongs to the septate division the name Derbya may safely be confined to this type of structure.

---

While I think there can be but little question that the shells from which the first description of the genus Orthotetes were drawn belong to the group for which Waagen's term Derbya is now in use, there might possibly be a difference of opinion as to the date which should be given for the establishment of the genus. Fischer de Waldheim's first description is not very clear, but it certainly implies structures the possession of which warrants the discrimination of Orthotetes from the other strophomenids. On the other hand, no figures accompany it, nor any list of species, one of which might be taken as a type. Figures were given in the Oryctographie (edition of 1830), but only such as represent the interior. No specific name was connected with the figures, in which are no intrinsic means for specific identification. The later edition, in 1837, contained a figure of the exterior of a dorsal valve, but again without a specific name or adequate means of identification with a known species. It was apparently these figures which were identified by De Verneuil (not D'Orbigny, as stated by Fischer, nor Bronn, as stated by Dall) with Orthotetes arachnoides, which Davidson regards as merely a variety of Streptorhynchus crenistria Phillips. This opinion has received general acceptance, and in this way the type species of Orthotetes is sometimes cited as Orthis crenistria, e. g., Hall and Clarke. Waagen's interpretation of Orthotetes was also determined by this species, and by the figures of the interior of Streptorhynchus crenistria which Davidson has furnished. He remarks: "There can be no doubt that Strept. crenistria, Phill., belongs to this genus, even if it should be proved that Fischer v. Waldheim's original specimen did not belong to the same species."a Streptorhynchus crenistria may have had no septum in the ventral valve, as it is described by Waagen and Davidson, or it may have had two septa (dental plates), as Schellwien represents it; but apparently it did not have a single median septum on any count. Thus it can not belong to the same species, nor, according to present ideas, to the same genus, as Orthotetes radiatus.

---

aWaagen, W., Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 607.

---

With the publication of 1850 we first have a thoroughly intelligible description of Orthotetes associated with the description of a species which might be taken for a genotype. Some might take the ground that the genus was adequately founded only at this time, the generic name being bound up with O. radiatus as its type. Were this all, I would even yet incline to recur to the earlier description of 1829, though a type was not associated with it until later, because that description does, after all, indicate the essential points, and clearly is in agreement with the 1850 description. The probability, amounting almost to certainty, however, that the specimens figured in 1830 were the same as those on which O. radiatus was based in 1850, and on which the generic description of that date depends, seems to strengthen the argument for going back to the earlier date.

Some argument, too, might be made in favor of retaining Waagen's name Orthothetes, even if Orthotetes does replace Derbya, on the ground that the two words are not the same, letter for letter; but they are near enough to occasion much confusion, in view of the close biological relationship of the two genera and the association of the names in taxonomy. I am also disposed to attach importance to the common sense argument that Waagen was not proposing a new genus in using Orthothetes, but was clearly redefining a group to which he believed Fischer de Waldheim had half a century earlier applied the name Orthotetes.

Reference has already been made to the uncertainty about the internal structures of Streptorhynchus crenistria. The interiors figured by Davidson can probably be taken as belonging to Phillips's species. If so, they can certainly be placed in the genus Schuchertella. Schellwien, however, figures specimens identified as Orthothetes crenistria, which have two long septiform dental plates. It is evident that these shells do not belong to the same species, and probably not to the same genus as those figured by Davidson and embraced in Waagen's definition of "Orthothetes." Apparently Schellwien held the same view, for he excludes from this group of biseptate shells, for which he employs the name "Orthothetes," a species which Waagen ascertained to be without ventral plates, and so to belong to "Orthothetes" as understood by him. This form, O. subplanus, Schellwien places in Streptorhynchus, apparently disregarding the difference in other respects which distinguish Streptorhynchus from Waagen's "Orthothetes," unless, indeed, he ascertained characters which showed Waagen's reference to have been an error. For the later representatives of this biseptate group, which differ from the earlier ones in having the dental plates parallel or converging, instead of very strongly diverging, Schellwien suggests the term Orthothetina. It is clear to me that these biseptate shells can not be placed in any of the subdivisions previously recognized; and I also doubt the advisability of distinguishing the earlier from the later types, unless other characters can be assigned than difference in the direction of the dental plates. Though it seems probable that these differences exist and will yet be ascertained, it is deemed best for the present to extend the term Orthothetina so as to embrace both forms.

The genus Orthotetes, its known representation the world over being considered, is much rarer than Derbya. Waagen found it to be lacking in the faunas of the Salt Range, while, strangely enough, Schellwien observed neither genus in the Alpine Permian fauna. The forms described by Abich from Armenia, which Waagen conceived to belong to Orthotetes, Schellwien has stated to be representatives of Orthothetina. The South American species Orthotetes correanus is a classical example of this group, and it is almost certain that the genotype, O. radiatus of Russia, was one of the camerates. Under the title of Orthis crenistria Trautschold has figured a Russian species which appears to have the camerate structure. A better illustration of the same specimen by Schellwien, however, makes it very doubtful whether it is not one of the septati. The interpretation of the figure depends largely on the inclination at which the view is supposed to be taken. If the view is foreshortened the structure may well be the septate type. But the outline of the specimen as a whole indicates that there is no foreshortening, and if the view is perpendicularly downward the structure would appear to be of the camerate type. In that case it may belong to Orthotetes radiatus, the type of Orthotetes. Schellwien's remark, however, that this specimen "zeigte die Merkmale von Derbyia in vortrefflicher Weise und dûrfte mit den hier in Text abgebildeten Exemplaren von der Indiga specifisch übereinstimmen,"a seems to settle the matter, for the text figure referred to clearly represents one of the septati.

---

aNeues Jahrbuch, Jahrg. 1900, vol. 1, 1900, p. 10.

---

In North America species possessing the camerate type of structure have not as yet been noted. Such as occur are probably rare, or possess the characteristic septal arrangement in an inconspicuous degree. The latter is true of Orthotetes keokuk, which seems to belong to this group. No post-Mississippian representatives of Orthotetes, however, are known to me, with the exception of the Guadalupian species described hereafter. Of these the forms discriminated are five in number—O. guadalupensis, O. declivis, O. distortus, O. distortus var. campanulatus, and Orthotetes sp. a. The four species named first are all found in the Capitan limestone, and show the characteristic camerate structure in a high degree of development, so that, with the exception of Orthotetes sp. a, whose generic affinities are doubtful and whose horizon is the basal black limestone, the development of this genus appears to have taken place in the higher measures of the Guadalupian section.

The Guadalupian representatives of Orthotetes, and also of Derbya, are distinguished from the Pennsylvanian species, and I suspect from many of those of foreign areas as well, by their extremely fine liration. One species (O. guadalupensis) is peculiar in having the liræ smooth, flattened, and crowded, instead of sharp, crenulated, and spaced, a type of surface which probably belongs also to O. declivis and differs from anything I have before seen in these shells. All things considered, this type constitutes a rather individual feature of the American fauna.

O. guadalupensis and O. declivis may be taken to represent one group which is distinguished from another, composed of O. distortus and O. distortus var. campanulatus, by having broad, rounded liræ, instead of thin, sharp ones, though the distinction is not easily enforced, because exfoliation is apt considerably to obscure this feature. These two groups are distinguished from Orthotetes sp. a, by having a much more elevated ventral valve.

ORTHOTETES GUADALUPENSIS n. sp.

Pl. X, figs. 1 to 5.

1859. Orthisina, sp. (?). Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 395 (data of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

Shell small, planoconvex, subpyramidal. Dorsal valve transverse, subcircular, or subelliptical. Hinge line straight, much narrower than the width below. Anterior margin straightened; lateral margins strongly and evenly rounded, merging into the anterior, but joining the cardinal margin with a faint angulation. Convexity moderate. Umbo rather prominent. Beak small, not projecting beyond the hinge line. Cardinal area apparently linear. Dental lamellæ long and not strongly diverging, making an angle of less than 90° between them.

Ventral valve high-conical. Area high, well defined, flat, divided by the delthyrium into two equal portions, each of which is again divided by a diagonal line running from the apex to the cardinal margin and situated somewhat farther from the outer edge of the area than from the edge of the delthyrium. The inner portion of each side of the delthyrium is somewhat raised and indistinctly striated longitudinally. The outer portion on either side bears conspicuous transverse striæ. The delthyrium is narrow and covered by a rather strongly convex pseudodeltidium, which extends almost to the cardinal line and is distinctly, though finely, cross striated. Longitudinally the area is sometimes flat, but more often somewhat concave. Occasionally the curvature is considerable, and sometimes the apex is more or less bent to one side or the other. The area usually seems to make a slightly obtuse angle with the plane on which the edges of the two valves unite. The expansion of the shell was, as a rule, symmetrical and regular, but rugosities now and then occur. Usually the line from apex to edge of the shell is nearly straight, occasionally somewhat concave, and when the apex is distinctly bowed is somewhat convex. This valve often has a faint sinus and the other an indistinct fold; but this feature is hardly to be perceived, except on the joined edges of the two valves, where it produces a faint sinuosity.

Surface marked by fine, regular, equal, radiating liræ, of which 16 or 17 are found in the space of 5 mm. Toward the hinge line they are somewhat curved. Increase is by implantation; and where some of the liræ are small and young the number given above is considerably increased. The liræ are depressed, rounded, not separated by relatively large intervals; neither are they alternating nor crenulate, as in so many strophomenoids, but rather resemble those of Orthis. This character, together with their fineness, makes this shell noticeable among the related forms in our collection.

Up to the present this shell is practically unknown save at the El Capitan locality (station 2926), where, however, it is abundant. A single specimen from station 2906, southwest of Guadalupe Peak, agrees closely with dorsal valves from the typical locality, and though in size it somewhat exceeds the types, a large ventral valve from the same station indicates about equal dimensions.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906); Delaware Mountain formation, Guadalupe Point (station 2963?), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2962? and 2969?).

ORTHOTETES DECLIVIS n. sp.

Pl. X, figs. 6 to 8a.

The following description is taken from the type specimens—a ventral and a dorsal valve found in the Capitan formation of the Guadalupe section. Variations shown by other specimens will be described thereafter.

Ventral valve high, conical. Area flat, moderately high, probably strongly inclined backward. Width of area at least 30mm.; height 13 mm. Pseudodeltidium about 5-1/2 mm. wide at base, rather convex, marked by fine transverse lines. Other characters as in Orthotetes guadalupensins. Shape probably transversely semicircular, more or less strongly contracting at the hinge. A line drawn on the surface from the apex to any point in the periphery would be nearly rectilinear, very slightly convex. Surface as in O. guadalupensis, with very fine flattened liræ separated by narrow grooves.

Three dorsal valves associated with the ventral described above have been referred to the same species. None of them is sufficiently perfect to show the shape, which is probably nearly semicircular, contracted at the hinge. The convexity is considerable; the beak small, depressed, and projecting but little beyond the cardinal line. Surface as in O. guadalupensins. The cardinal process is long and bilobed, and its direction is nearly at right angles with the plane of the shell edge. The dental lamellæ are long and mutually directed at an angle of about 60°.

In the foothills southwest of Guadalupe Peak, where the Capitan limestone is faulted down to a much lower level, a form occurs in considerable abundance, though imperfectly preserved, which seems to be conspecific with the types. The ventral valve in some specimens is proportionately a little narrower and higher; in others a little broader and lower. The type specimen is peculiarly regular and symmetrical; these others are more or less distorted and uneven. It is impossible to ascertain the real shape of these specimens, because the anterior portion is in every case broken away, but the dorsal valves appear to have normal proportions. The dorsal valves with which these ventrals are associated agree with those from the main range, except that the dental plates are less lengthy than in the somewhat extreme example figured. A dorsal from this locality, of the usual type, is represented in Pl. X, fig. 8. The type ventral valve does not show the character of the septum and dental plates, but others demonstrate the presence of a large chamber.

The most nearly related species is evidently Orthotetes guadalupensis, which occurs in the same beds. From this form O. declivis is distinguished in several ways. It is a much larger shell, and the ventral valve has a lower and broader area. The shape does not contract as rapidly near the area as in the smaller form. The dorsal valve closely resembles that of O. guadalupensis, but is larger and has proportionately longer dental lamellæ.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906), Guadalupe Mountains, Texas.

ORTHOTETES DISTORTUS n. sp.

Pl. X, figs. 9 to 9c.

Of this species but a single specimen has been obtained, and it presents differences from Orthotetes guadalupensis, with which it is associated, so marked that they can hardly be referred to the same specific group. O. distortus is a small shell, much inferior in size to O. guadalupensins. The maximum width is only 11 mm. Both valves are highly unsymmetrical and distorted. The hinge line is narrower than the shell in front. The dorsal valve is moderately convex; the ventral is relatively high, measuring about 6 mm. The area, about 6 mm. in width, is flat, and the pseudodeltidium convex, but broad, low, and not well defined. In this regard it is in marked contrast to O. guadalupensins. It also bears a longitudinal callosity that is narrow and almost linear. The surface is marked by extremely fine subequal liræ, of which 35 or more occur in the space of 5 mm. The liration is thus seen to be much finer than in O. guadalupensis, and is, in fact, but little more than just visible to the naked eye.

In my preliminary list I referred this shell to Derbya bennetti?. The general aspect is rather similar to that species, but the Guadalupian form is much smaller and more finely striated, besides possessing the structural peculiarities of the camerate group.

To this species has provisionally been assigned a small, fragmentary, silicified example from the Glass Mountains. It is distinguished primarily by its extremely fine liration, in which it agrees with O. distortus, itself unique among Guadalupian strophomenoids in this particular. The liræ, however, are a little thinner, with relatively wider striæ between. The hinge plate is large, with a tendency to extend downward into the valve as if when fully grown it might have surrounded a muscular area. The fragment has a length of scarcely 5 mm., but this appears to be nearly the original dimensions.

It is of course highly uncertain whether this specimen belongs with the Capitan species. No other identification, however, appears equally promising.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

ORTHOTETES DISTORTUS var. CAMPANULATUS n. var.

Pl. X, figs. 10 to 10c.

Of this form a single specimen has been obtained—a ventral valve. It resembles Orthotetes distortus in a general way, and yet presents some marked differences, and being uncertain whether it would with propriety be thrown with that species, on the one hand, or made the type of a new one on the other, I have chosen what may be considered as a middle course and distinguished it from O. distortus merely as a variety. Only the ventral valve of this species is known, and from it the following description has been drawn up:

Shell rather small, high, conical, campanulate. Growth irregular and asymmetrical. Area narrow, high, nearly flat, but twisted and slightly concave. It is poorly defined from the convex walls of the shell. Its general direction to the plane of the shell edge is nearly vertical or slightly inclined forward. Hinge line 7 mm., much shorter than the width in front, which is 16 mm. Length of aperture 17 mm. Nearly all of the area (about 5 mm.) is occupied by the pseudodeltidium, which is convex, and by the greatly thickened edges of the dental plates. The latter are rather long, uniting with the septum, which extends well nigh to the edge of the shell. The liræ are very fine, equal, regular, subangular, and depressed, separated by intervals about equal to their own thickness.

In its irregular growth and rapid expansion near the aperture this shell develops several distinct plications, which might possibly warrant the assignment of it to the genus Geyerella. It appears, however, that this is but an accidental feature. The larger size, greater relative height, and narrower area should distinguish this form from O. distortus.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

ORTHOTETES? sp. a.

This division is introduced for a type occurring in the black limestone which forms the basal member of the Guadalupian series. Only a single ventral valve has been found, and on it the following description has been based:

The area is low, flat, and either at right angles to the plane of the edge or slightly inclined forward. Width about 12 mm., height about 3 mm. The pseudodeltidium is moderately convex, about 3 mm. wide at the base. The shape of the shell is depressed, subconical. The growth is symmetrical, but somewhat irregular. The maximum width is about 18 mm., the length uncertain. The liræ number about 20 in a space of 5 mm., the measurement of necessity being taken not far from the apex. They are slender, moderately high, and separated by intervals greater than themselves. Fine concentric liræ of considerable strength can also be noted, and at present are confined to the flat interliral spaces. The septum is long and thin, and probably unites with the dental lamellæ. It is seen to be continuous with the plate on one side, but that on the other side appears to be missing.

As it is not clear to which of the two great groups of Orthotetes (Orthotetes sensu stricto or Derbya) this form belongs, it seems hopeless to attempt to fix its specific position. Its most noticeable features are its low altitude, small size, and fine, sharp liræ. The latter feature at once distinguishes it from the group of Orthotetes guadalupensins. It is in fact more similar in superficial appearance, and possibly in internal structure, to Derbya sp. a, but appears to differ in having more regular growth and a relatively lower and more upright cardinal area, and besides being much smaller, it has finer liræ and occurs at a lower geologic horizon.

To this same group I have provisionally assigned an imperfect dorsal valve from the Glass Mountains, which is distinguished by having very fine, slender, somewhat spaced, and occasionally alternating liræ, long, straight socket plates, and a long cardinal process directed nearly at right angles to the plane of the edge. The size attained was upward of 20 mm. in transverse diameter. The sculpture in this specimen is largely obscured, but appears to be very similar to the ventral valve from the black limestone, though possibly somewhat finer. The prolonged and strongly elevated socket plates are suggestive of the form described as Orthotetes declivis. The chief positive difference resides in the liræ, which are broadly rounded in declivis, though they appear more angular when exfoliated, their usual condition. As the Glass Mountain specimen is silicified, however, this explanation can not be invoked in order to unite them. Except for this circumstance I would have tentatively placed this shell with O. declivis.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2920). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Genus GEYERELLA Schellwien.

Geyerella is another name introduced by Schellwien, and is valid for the group to which it was applied, though hardly, it appears to me, in a full generic sense. The shells which it includes are especially distinguished by their plicated surface and by the internal structures of the ventral valve, where the long dental plates converge and unite with the median septum. This is precisely the internal structure of the camerate group of Derbya, a fact to which Schellwien does not, in my view, give sufficient weight—indeed, I do not know that he calls attention to it at all. Geyerella then stands in the same relation to Orthotetes that Meekella does to Orthothetina, and that the plicated group of Streptorhynchus does to the simple one. I can not accord these plicated shells full generic rank, nor am I willing to reduce them to mere divisions. In any event, it appears to me that all should be treated alike.

The only American representative of this genus at present known to me is the species described below. It apparently does not possess any marked traits by which it is strongly distinguished from the European species.

GEYERELLA AMERICANA n. sp.a

Pl. XI, figs. 2 to 2b.

---

aIn the preliminary notice of this species I referred to it as Meekella n. sp.

---

This species is represented in our collections by a number of specimens, all of which, unfortunately, are extremely fragmentary. The most complete example is represented by the figures, and even that is better for showing generic characters than specific. It consists of part of the ventral valve of a large shell, the dorsal valve of which is missing. The height when complete must have been about 40 mm.; the length from the anterior margin to the center of the area is also about 40 mm. The shape is thus seen to be rather elongate conical. The area makes nearly a right angle, possibly a somewhat acute angle, with the opening of the shell. The delthyrium is rather broad, and the convex pseudodeltidium is probably somewhat depressed below the surface of the area.

The surface is marked by about 26 strong but irregular plications, which become obsolete on both sides near the area. A few of the plications bifurcate. The superficial liræ with which the surface is covered are almost obliterated by exfoliation; but they appear to be fine and equal, and manifest a tendency to curve upward from the sides of the plication onto their crests, just as in the genus Meekella.

Geyerella americana finds its closest ally in G. gemmellaroi; but as Schellwien figures that species b without, so far as I have been able to discover, describing it, and especially as I have not been able to examine any specimens belonging to it, I find it impossible to make a close comparison. The present species shows more strongly marked plications than G. distorta or G.? eusarkos.

---

bNeues Jahrbuch, Jahrg. 1900, vol. 1, Separat-Abdruck, 1899, pl. 1, figs. 7a, 7b.

---

Shumard cites from Guadalupe Peak a form which he calls Streptorhynchus (Orthisina) shumardianum, a name now generally regarded as a synonym of Meekella striaticostata. It is possible that the form mentioned may have been a representative of this species, or it may have been a true Meekella, though not probably one conspecific with Cox's species.

This type of structure and surface combined is entirely unknown in the faunas of the Mississippi Valley, and this first citation of Geyerella from either of the Americas is attended with some interest.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

Genus MEEKELLA White and St. John.

This genus is distinguished from Orthothetina very much as Geyerella is distinguished from Orthotetes, and both groups can hardly be allowed full generic rank. The distinguishing features of Meekella are its plicated surface and its long, nearly parallel dental plates. A remarkably well-preserved example in the Guadalupian fauna possesses structures compelling a doubt, however, as to whether these septa are really dental plates. They do not meet the area along the margins of the delthyrium, but considerably to the sides, while from the delthyrial margins, where the dental plates should originate, a pair of small platelike structures arise, bend outward, and unite with the two septa. (See fig. 10, Pl. XXX.) Indications of a similar structure have been noted in Meekella striaticostata. What interpretation should be placed on these observations is doubtful, but they may mean that the two septal plates correspond to the single septum of Derbya and Orthotetes, and that the real dental plates join but are not continuous with them.

It is perhaps rather striking that this genus has been found only in the lowest beds of the Guadalupe section. It is true, however, that Shumard cites Streptorhynchus shumardianum Swallow (= Meekella striaticostata Cox) from the sandstones near the base of the Guadalupe Mountains. Whether this was a Geyerella or a Meekella will probably never be known. From the black limestone at the base of the section we have M. attenuata and M. multilirata.

The three other species distinguished in this paper were obtained in other areas, but probably from the horizon of the Delaware Mountain sandstone, and it may be one of them that Shumard found in the Guadalupe Mountains. These four species of Meekella are rather closely related to one another, and do not lend themselves readily to separation into different groups. Perhaps the best subordinate classification would assemble Meekella attenuata, M. skenoides, and M. multilirata in one group, and leave M. difficilis by itself to represent another.

MEEKELLA ATTENUATA n. sp.

Pl. XXIV, figs. 7 to 9a;>BR> Pl. XXV, figs. 4 to 4d.

Shell small. Ventral valve high, subconical, inclined backward, so that the area makes an angle of not far from 135° with the plane of the shell opening. Area well defined, generally plane, but more or less contorted; width about 7 mm., height about 6-1/2 mm. Pseudodeltidium apparently rather broad, nearly flat, and not well defined. The growth of this valve is marked by constrictions and contortions.

Dorsal valve subcircular, outline regularly rounded, with somewhat straightened and converging sides. Width of hinge 7 mm.; width below 10 mm.; length 10 mm. Convexity moderate. Growth rather unequal and irregular.

Surface marked by nine or ten plications, which are low and not very distinct, reaching but a short distance back from the anterior margin and not extending to the sides. The liræ, which furnish the only ornamentation over much of the shell, are, for its size, comparatively coarse. They are more or less alternating, but tend to become finer over the plicated portion of the valves.

This species is distinguished from either Meekella skenoides or M. difficilis by the coarseness of its liræ and the faintness of its plications. The separation recorded here is that which seems to be demanded by the material studied, but this is scanty and in some instances incomplete. The relationship to one another of these three species is more or less close, and owing to the well-known variability of shells belonging to Meekella it is possible that some of the differences, though now rather striking, will be bridged over by intermediate forms.

In the basal black limestone of the Guadalupe section (station 2967) occurs in abundance, though in a poor state of preservation, a species of Meekella which with our present knowledge it seems unavoidable to refer to M. attenuata, although necessitating some minor change in the limitation of the species. This change is connected with the configuration, in regard to which Meekellas are never very constant. The specimens from the black limestone attain a size considerably greater than the specimen described, and exhibit variations, both in the elevation of the ventral valve and the angle at which the area extends to the plane defined by the margin of the shell. On the other hand, the black-limestone specimens agree in having the liration equally coarse, in having the earlier stages unplicated, and in having the plications, when they do appear, relatively faint and fine. It is unfortunate that the specimen from the Diablo Mountains, from which the description is taken, is not accompanied by enough material to show the range in size and other characters which this species manifests.

Comparisons of the typical specimen with the common M. striaticostata of the Mississippi Valley are scarcely necessary, but the larger, more mature examples from the black limestone present a much closer resemblance. The two species should be readily discriminated by their sculpture, for the liræ in M. attenuata are rounded, crowded, and coarser than in M. striaticostata, in which they are thin and separated by relatively wide striæ. M. pyramidalis Newberry and M. occidentalis Newberry are too imperfectly known to make comparisons of much value. They are much larger species than those described here, and probably from a different geologic horizon. It is interesting to note that Waagen describes no species belonging to this genus from India, nor does Abich figure any from Armenia. In the Carnic Alps, however, Schellwien found a number of forms, some of them of large size. None of these seems specifically identical with those described here, nor have I the material at hand for adequate comparison, except in the case of M. irregularis, which seems to be distinct.

Horizon and locality.—Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764). Basal black limestone, Guadalupe Point, Guadalupe. Mountains, Texas (station 2967).

MEEKELLA SKENOIDES n. sp.

Pl. XXX, figs. 8 to 9.

Shell small. Ventral valve high, symmetrical, subconical, leaning backward, so that the area makes an obtuse angle with the plane of junction of the valves. Area well defined, high, flat transversely, slightly concave longitudinally; width 8 mm., height 9 mm. Delthyrium of medium size, about 2 mm. in width at its base. A line following the shell from apex to anterior margin would be gently convex; to the lateral margins it would be slightly concave.

Dorsal valve transverse, subcircular. Outline regularly rounded below, contracting at the hinge line to 8 mm.; width below, 13 mm.; length 11 mm. Convexity moderate, with umbo rather gibbous.

Surface marked by 11 or 12 plications, which are high and thin over most of the shell, but die out as the cardinal area is approached. The superficial liræ are fine, subequal, and persistent over the whole surface. They appear to be more or less interrupted or nodose, but this may result, in part at least, from the coarse silicification which the shell substance has undergone.

The foregoing description is based on the type specimen alone, which can readily be distinguished from the type of Meekella difficilis by reason of its smaller size and more lofty ventral valve, making a proportionately narrower and higher area, and by the persistence of the striation over the whole surface. To the same species, however, I have referred certain other shells, for the most part more or less fragmentary, which if actually conspecific, as they appear to be, introduce certain modifications of the characters above set down. These are chiefly dorsals, and indicate a width of 20 mm., or even greater in one instance. All, however, are conspicuously lirated, the liræ toward the front curving upward from the bottom of the sulci. One small ventral valve has a lower and broader area, with the apex bent to one side.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763). Delaware Mountain formation, southern Delaware Mountains, Texas (station 3501).

MEEKELLA DIFFICILIS n. sp.

Pl. XXX, figs. 10 to 10g.

Shell of medium size. Ventral valve subconical; height about 13 mm., width. 24 mm. Area well defined; flat in a transverse direction; slightly concave longitudinally. Pseudodeltidium narrow, slightly convex, poorly defined. An indistinct line dividing the lateral portions of the area about midway indicates the attachment of the dental lamellæ. These structures converge toward the shell wall, which they join in nearly parallel lines, somewhat diverging anteriorly and reaching a little less than halfway from the apex to the front edge. Two other shelly plates, originating one on each margin of the delthyrium, bend strongly to the sides, so that they are nearly parallel to the plane of the area, and become adnate with the two septa just before they join the latter. It seems as if we should call these the true dental plates and the others independent structures. If the large septa are really the dental plates and these small laminæ merely auxiliary, the relation of the latter to the pseudodeltidium and of the pseudodeltidium to the whole area is hard to account for. Surface from beak to front somewhat convex; from apex to sides slightly concave. Cardinal process long and four pronged.

Dorsal valve moderately convex; shape transverse, broadly rounded, somewhat contracted toward the hinge line, which is a little narrower than the shell below.

Surface marked by 11 or 12 strong angular folds, which become less and less distinct laterally. Toward the apex also the plications subside and resolve into moderately coarse subequal liræ. An intermediate stage exists in which the bottoms of the sulci are occupied by liræ considerably smaller than the incipient ridges. Over most of the shell, except at the sides, where the plications are obsolete, the liræ appear to be completely absent, though they may be somewhat obscured by the coarse silicification which the specimen has undergone.

The development of the dental plates in this species, to which reference has already been made, is peculiar. As in most strophemenoids, the area is intersected on each side by an oblique line, extending from the apex to the hinge, which divides more or less equally the triangular surface defined by the edge of the area, the hinge line, and the delthyrium. The two areas on each side of the delthyrium are symmetrical and marked by hatchings extending in one case longitudinally and in the other transversely. In the present species the two plates are directed, not toward the margin of the delthyrium, from which at the hinge line the cardinal teeth presumably project, but to the oblique intersecting line. From the edge of the delthyrium, however, two other plates originate, one on each side, which diverge strongly from one another at an acute angle with the area wall, and become consolidated with the septa not far from where they meet the latter. The interpretation of this singular construction is not entirely obvious, but it may indicate that, to speak strictly, the septa are not dental plates but independent structures.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains (station 3763), and mountains northwest of Marathon (station 3840), Texas.

MEEKELLA MULTILIRATA n. sp.

Pl. XXIV, figs. 6 to 6b.

Shell rather small. Length from back to front 21 mm., width 25 mm. Height of ventral valve 12 mm.

The shape of the ventral valve is subcircular at the aperture, much contracted at the hinge line, which has a width of but 14 mm. The area, which is not very sharply defined, is slightly concave longitudinally and strongly inclined backward. The delthyrium occupies about one-third of its width, being 5 mm. wide at its base and 13 mm. high.

The surface is marked by 10 or 11 moderately strong, simple, rounded plications and by very fine, rounded, radiating liræ. The latter have been largely exfoliated, but where preserved come about 6 in the space of 1 mm. They run up from the bottom of the sulci on the sides of the plications in the manner often found in and to a considerable extent distinctive of this genus.

This species is based on the ventral valve represented by my figures, the dorsal valve being unknown. In discriminating this species reliance has been placed not so much on the configuration as on the very fine liration. It occurs associated with shells identified as Meekella attenuata, and resembles them in shape rather closely, but is somewhat larger, and above all is much more finely striated. In a general way this shell resembles M. difficilis; it differs, however, not only in the finer features of its surface but in having the plication and sulci lower and much less angular. It is more nearly related to M. striaticostata than the associated shells referred to

M. attenuata, and I am not satisfied that it will not prove to be identical with our well-known Pennsylvanian species. At present it seems to be distinguished by its less well-defined area, wider pseudodeltidium, and less strongly plicated surface—characters all liable to modification by exfoliation, to which the Guadalupian specimen has been subjected—and to some extent by the sculpture, though this also has been largely lost by exfoliation. The liræ are possibly no finer than in characteristic M. striaticostata, but they appear to be low, rounded, and closely arranged. In the Pennsylvanian species they tend to be thin, spaced, and more or less alternating.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2967).

Genus ORTHOTHETINA Schellwien.

Schellwien's interpretation of the genus Orthotetes so far differs from Waagen's, and indeed from that of the original author, that he employs the name for shells having two elongated dental plates like septa in the ventral valve. Waagen, equally wide of the author's real meaning, uses it for shells without any partitions whatsoever. Both—erroneously, it may be said—employ for the type species Streptorhynchus crenistria Phillips; but Schellwien's material was specifically—I can not but think generically—distinct from Phillips's species, while Davidson's specimens, from whose illustrations Waagen's conception of the structure of S. crenistria was doubtless in part derived, are more probably correctly identified. Schellwien found that in the upper Carboniferous and Permian forms which have this structure the dental plates are parallel and close together, while in the Devonian and early Carboniferous forms they are strongly diverging. For the former Schellwien suggested the term Orthothetina.a The manner in which this name was put forward was apparently merely a suggestion, the first use being found on pages 8 and 13 of Beitrage zur Systematik der Strophomeniden des oberen Palaeozoicum,b and later he appears to have abandoned it altogether.c

---

aThe form in which the author wrote this name was Orthothetina, but as the name is evidently derived from Orthothetes it seems that it should be changed to Orthotetina, in accordance with the corrected spelling of that name. If written Orthothetina, however, it is less liable to be confused with Waagen's subfamily term Othotetinæ.

bNeues Jahrbuch, Jahrg. 1900, vol. 1, Separat-Abdruck, 1890.

cAbbandl. K-k. geol. Reichsanstalt, vol. 16, heft 1, 1900, p. 16.

---

I have attempted to show on page 198 that Streptorhynchus crenistria can not be employed as the type of Orthotetes, and that the name itself can not be applied either to the group for which Waagen or to that for which Schellwien used it. For Orthothetes Waagen (non Fischer de Waldheim) I have proposed the name Schuchertella. For Orthothetes Schellwien (non Fischer de Waldheim) Schellwien's name Orthothetina, though by him applied only to a portion, may be used. If it should prove, on further investigation, that the late Carboniferous species must be separated from the Devonian and early Carboniferous ones, it is to the former that the name Orthothetia will attach. No American representatives of Orthothetina are known, though I have one or two undescribed forms in addition to the one whose description is given below without a name. This seems to be a veritable Orthothetina, but it is so imperfectly preserved that but little can be said as to its specific relationship.

The internal structures of this group are precisely those of Meekella, so far as known, the only ascertained difference being that it has a smooth shell instead of a plicated one. This difference, is hardly generic in degree, but the two names can probably be retained, one as a subgenus of the other. It might be thought that Meekella was a later development of Orthothetina, but the occurrence of the two genera in America appears to refute this supposition, and to indicate that they originated independently, or that Orthothetina was a retrograded form of Meekella, for the latter genus appears early in our upper Carboniferous series, while the Orthothetinas, so far as known, came in somewhat later, though while Meekella still survived.

ORTHOTHETINA sp.

The shell on which this type is based is a ventral valve found, along with Orthotetes guadalupensis, on the east side of El Capitan (station 2926), in the Guadalupe Mountains. It is a type less small and delicate in construction than its associates. Unfortunately, its condition is now too imperfect to permit a full description and a discussion of its specific affinities. It appears to have been semi-circular in shape, probably somewhat transverse, with the hinge line nearly as wide as the shell below. The area appears to have been high, generally flat, though considerably warped, and rather strongly inclined backward. Its width was probably about 40 mm. and its height not much less than 25 mm. The maximum width of the pseudodeltidium appears to have been about 10 mm., a good portion of which, however, was occupied by greatly thickened dental lamellæ. These structures are long and converging, and they join the vaulted anterior wall just before the point of their union with one another. There is thus no septum in this valve. The surface ornamentation has been largely lost through exfoliation, but appears to consist of slender, spaced, radiating liræ.

This genus has not previously been recognized in America, and it is to be regretted that the present example is too imperfect for illustration.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

Family THECIDEIDÆ Gray.

Subfamily LYTTONIINÆ Waagen.

When Waagen described the subfamily Lyttoniinæ but two generic groups were known to belong to it—Leptodus (Lyttonia) and Oldhamina. Tschernyschew has recently introduced a new generic name (Keyserlifingina) for a smaller and simpler type, which nevertheless without much doubt belongs to the same subfamily. For many years the Lyttoniinæ were, so far as known, restricted in their distribution to eastern Asia, Leptodus being found in India, at Timor in the Indian Archipelago, and in China, and Oldhamina only in India. Keyserlingina has at present been recognized in the Carnic Alps and in Russia. It is of some interest, therefore, to note the occurrence of this peculiar brachiopod type in the United States, and it should be remarked that the forms which occur here distinctly resemble the Asiatic rather than the European group of species, and lend a decidedly Asiatic or non-Pennsylvanian aspect to the Guadalupian fauna.

Although for a long time Leptodus appeared to be restricted in distribution to southeastern Asia, the occurrence in the Guadalupe Mountains is not the first instance that is known of its appearance far from the region of its discovery, for I learn from Diener that Gemmellaro has determined the genus in his fauna from Palermo. Unfortunately, I have been unable to obtain a copy of Gemmellaro's paper, and consequently can do no more than note the fact, of which there are many of a similar sort, that the Sicilian and the Guadalupian faunas possess this singular type in common.

I have cited both Oldhamina and Leptodus from the Guadalupian, and while the presence here of the former type is a matter of some doubt so far as my material is concerned, it can be said with assurance regarding the latter that the two species which have been recognized are congeneric with Waagen's Lyttonias. Specifically there can be no question that both are distinguished by important differences from the two Indian species. It seems probable that they are more closely related to the Chinese form L. richthofeni, but the latter is so imperfectly known that it is impossible to determine how close the relationship really is.

So far as the Lyttoniinæ are concerned, therefore, the Guadalupian fauna may be said to be related to those of the Salt Range, of the Himalaya,a of China, and of Palermo, rather than to those of the Carnic Alps and of Russia; and if, as Professor Schuchert has suggested to me, the relative complexity of the septal ridges may be used as some sort of an index of chronologic sequence, the Guadalupian fauna, so far as this type is concerned, may be younger than the Gschelian stage of the Russian fauna.

---

aIn the Himalaya Leptodus baa been cited by Biener from Kasbmir, from "Chitichun No. 1," and from Malla Sangcha the species being, so far as can be told, similar to those of the Salt Range.

---

Tentatively, in view of the their specific characters, the Guadalupian forms are possibly more closely related to the Lo Ping occurrence than to the species of the Salt Range and the Himalaya, the Sicilian type being for the time left out of account.

Genus LEPTODUS Kayser.

The generic name which should be employed for these shells is a matter of some uncertainty. Kayser in 1882, thinking that the fossil was a fish tooth, named it Leptodus. This name Waagen rejected, on the ground that it is inappropriate, and in 1887 he proposed instead the term Lyttonia, which has found general acceptance.

About the inappropriateness of the term Leptodus and the incorrectness of Kayser's conception in regard to it there can be little doubt. Yet this is not generally regarded as a valid argument for substituting a new name for one already proposed. On the other hand, though the name Leptodus in its literal form has not been preoccupied, we have Leptodes Dej. 1833, Leptodes Swains. 1839, Leptodon Raf. 1820, Leptodon Sund. 1835, and Leptodon Gaudry 1860. The first of the names mentioned was applied to a beetle, the second to a fish, the third to a mollusk, the fourth to a bird, and the last to a mammal.

As to its derivation, Leptodon has the same source as Leptodus. Leptodes Dej. is differently derived, while the derivation of Leptodes Swains, is not known to me. Leptodus and Leptodes are certainly different words, and in one case at least, as has just been seen, have different derivations. As they were proposed for very different groups of organisms there is little or no danger of confusion, and it seems that Kayser's name can safely be retained so far as Leptodes is concerned. Leptodon, though having the same origin as Leptodus, is still more different from the latter than is Leptodes, and like Leptodes, except in the case of Leptodon Rafinesque, applies to very different zoological groups. It seems to me, therefore, that none of the circumstances related can be considered as interfering with the validity of Leptodus Kayser, which is here adopted.

LEPTODUS AMERICANUS n. sp.

Pl. IV, figs. 8 to 8b; Pl. XXV, figs. 1 to 3a.

Shells belonging to Leptodus have been found in the Guadalupe, Glass, Diablo, and southern Delaware mountains. Those from the Diablo Mountains are silicified, and it has been possible to ascertain their characters with much greater completeness than in other collections, where the preservation is for the most part imperfect.

These specimens show a rather small species, more the size of L. tenuis Waagen than L. nobilis Waagen. Its greatest length can hardly have been greater than 50 mm. The largest fragments in our collection are only about half that. But little can be said in regard to the shape, which was probably narrow, truncated at the posterior end and expanded in front. As a rule ventral valves appear to have been more or less flat longitudinally and curved transversely, having what may be called an irregular scoop-shaped configuration. They were attached for a varying but small portion of their extent at the posterior end. The outer surface is marked by obscure transverse folds and by finer striæ. These do not conform altogether to the direction of the septa on the inner side, and if the anterior outline followed their direction it must have been broadly and gently curved.

On the inside there is a low, narrow median septum, with lateral septa. The latter are extremely variable, depending considerably on the distortion of the shell. They may be either nearly straight or strongly curved, the convex side directed toward the hinge; they may be either normal to the median septum or strongly inclined to it; they may be high or low, and are variously shaped. Usually they are thin and high at the inner end, becoming low and flattened at the outer, where they are crowned by a double row of pustules. In other specimens, or even other parts of the same specimen, they have a very different character, consisting of double ridges uniting into a loop at the inner end. This is perhaps the primitive form, later shell deposits elevating and unifying the constituent ridges except at the outer ends. Sometimes the median septum also is double. The shell was probably normally extended beyond the septate portion into broad lateral expansions more or less upturned. The entire inner surface is covered with papillæ, but these are more abundant and noticeable on the lateral expansions. They occur in double rows on the crests of the septa, as previously remarked.

The dorsal valve is small and pinnate in shape. Apparently the lateral branches do not unite into a solid rim. On one side (the inner) there is a median elevated ridge or septum, which is made double by a groove down the center. Farther forward the groove develops into a fissure, so that the valve becomes bilobate. The lateral branches are concave, with an elevated rim which follows the sinuous margin. The other (outer) side of the dorsal valve is distinguished principally by its strongly pustulose surface. It has a slight mesial groove and the different branches are convex. Waagen states that it is the outer side of the dorsal valve which is pustulose, and I have oriented my specimens accordingly. I would have thought, but for this, that the pustulose side was the inner one, since the inner surface of the other valve has this character. I have not observed the dorsal valve in place, however, and of the two small imperfect dorsals which have been observed, though both are nearly flat, one is more or less concave and the other slightly convex.

No specimen has come to hand from the "dark limestone" and only a single fragment from the sandstones of the Delaware Mountain formation. This fragment, preserved as an internal mold of the ventral valve, shows what would be a smooth central band, along the middle of which ran a median septum and at the sides of which began the lateral septa. These were simple, thin, and high, and the general appearance must have been that of the Diablo rather than that of the Guadalupe specimens. I have accordingly identified this form provisionally with L. americanus.

Horizon and locality.—Base of Capitan formation, hill southwest of Guadalupe Point (station 2906); Delaware Mountain formation, Guadalupe Point (station 2931), Guadalupe Mountains, Texas. Delaware Mountain formation, Diablo Mountains, Texas (station 3764). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

LEPTODUS GUADALUPENSIS n. sp.

Pl. IV, figs. 6 to 7.

In the white limestone of the Guadalupe Mountains Leptodus is fairly abundant, though few specimens are well preserved. The ventral valve usually presents itself as an internal mold, more or less of the shelly substance adhering to it. The dorsal valve is mostly very fragmentary. Unlike the form from the Diablo Mountains, specimens in this region as a rule occur singly, and apparently seldom grew in groups cemented to one another. The largest fragment observed measures 40 mm. and when entire probably reached 50 mm. or more. The width is much less, the shape being elongate, gently tapering toward the posterior end. Longitudinally the shell was nearly flat; transversely it was gently convex at the bottom, with the sides upturned and irregularly folded inward.

On the interior a median septum was present, at least in some examples. Seen in the reverse, as partial molds, the lateral septa present the following appearance: There is a series of rather slender raised ridges arranged pinnately to the median line, connected by curved continuations at alternate ends, so that the whole makes a more or less connected series of closely arranged sigmoid curves. The inner curves occur at short and regular distances from the mesial line, the outer near edge of the shell, where the upturned margin begins. These elevated ridges are studded with a single row of pustules, which are probably not altogether due to weathering, though they may have been exaggerated by it. Partly surrounding each of the lobes thus formed is a depressed line or channel. These pairs of pinnate channels are united at the outer but not at the inner end and thus make narrow U-shaped figures. The channels evidently correspond to the septa of the silicified specimen (of L. americanus), the elevated loops being the depressions by which they are separated. We have here, however, a distinct elevation in the center.

An example showing these characters is represented by fig. 6, Pl. IV, while fig. 6a is an artificial impression representing the real configuration of the interior of the same shell, except for the accidents of preservation. Here we see an elevated central band bearing a low median ridge or septum. Extending pinnately from the central area are loop-shaped grooves, the bottoms of which are conspicuously marked by rather large pits. The centers of the loops are gently elevated, and they are separated by high, rather thin, ridges, which are connected at their outer ends by less prominent curved continuations. Toward the front the structure at first appears to be very different, but a careful comparison shows that in fact it is constructed on the same plan, though with some modification. The diverse appearance seems to be due to an increased prominence of the area inclosed by the loop-shaped groove, so that it equals or dominates the ridges dividing the grooves from one another, which more posteriorly stood relatively higher.

This specimen has the sides much extended beyond the septate portion, rolled up, and irregularly arched over. It seems to me probable that this was the normal condition in this species. The slight protection to the animal parts afforded by the semiobsolete dorsal valve would apparently need to be augmented by these expansions of the ventral shell. Their apparent absence in many cases would be amply explained by the breakage due to their thinness and lack of support.

The specimen and impression on which the foregoing description is based are from the Capitan limestone at station 2926. Other specimens from the same station are more imperfect and less well preserved. The central idea of the lateral septa seems to be a loop-shaped ridge bounded within by a groove having the inclosed band gently elevated.

Several examples, somewhat differently preserved, have been obtained from the same horizon at station 2906. Two of these retain the shell, but are more or less exfoliated, so that their assignment to Leptodus is not confined to the evidence of the outside. They have a broadly triangular shape, flattened over the back, but turned up at the sides. Here again the lateral lobes of the septa consist of a loop-shaped ridge followed within by a groove surrounding a gently elevated central band. At this locality was found a specimen which with little doubt represents a very young example of the same species. It is a small, triangular, scoop-shaped shell, having a length of 6 mm. It is cemented at its posterior end to a branched bryozoan. The outside is somewhat irregular, but generally smooth. The inside is partially covered by the dorsal valve, which in this instance is four-lobed or four-rayed, being divided into two main branches, each of which again subdivides. The four branches fit into four depressions in the ventral valve, the remainder of which is unprotected. This specimen show the character of the articulating structure. The dorsal valve terminates in two lateral projections, which are surrounded by an inflection of the shell of the ventral valve, thus effecting a crude sort of hinge, one very different from the normal brachiopod structure.

It should be noted that the exterior of this valve is not papillose, but, on the other hand, the specimen is so immature that the exposed portions of the inner surface of the ventral valve are very slightly papillose, if at all.

From the Glass Mountains only two fragments have been examined. One of these shows a longitudinal median ridge, having indications of a faint septum. The lateral septa terminate at this with a strong backward bend, and their outer ends are connected by curved prolongations. Intermediate between two of these prominent ridges is another fainter one, which is nearer the proximal than the distal of the two septa between which it lies. The grooves separating the ridges seem to be marked by comparatively large pits.

I originally united all the occurrences of Leptodus into a single species, but became far from satisfied with the course. Apparently two groups can be distinguished—one represented by specimens from the Capitan formation of the Guadalupe Mountains and from the Glass Mountains, the other comprising specimens from the Delaware Mountain formation and from the Diablo Mountains. The septal lobes in the former appear to be less distinctly formed, to be somewhat more closely arranged, to lack pustulous ornamentation on the crests of the ridges, and to contain pits at the bottom of the grooves not found in the other. These characters are all connected with the septal ridges, but there are also other differences in the configuration and mode of growth found at least in the typical specimens. These forms manifest such variation in the character of the septa, however, not only in specimens from the same locality, but in the same individual, that I feel some doubt about trusting to the differences which appear to exist without more extended observations than my present material permits.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

Genus OLDHAMINA Waagen.

OLDHAMINA? sp.

Two fragmentary specimens from the Capitan formation of the Guadalupian are distinguished from Leptodus guadalupensis and Leptodus americanus, which are fairly constant in being nearly flat longitudinally, by having a strong curvature in that direction. Otherwise the characters appear to be essentially the same.

The inrolled shape, together with the massive hinge plate, discriminates Oldhamina from Leptodus. The hinge in my specimens has not been made out, but because they possess the marked longitudinal convexity they have been provisionally referred to Oldhamina.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

Family PRODUCTIDÆ Gray.

In most faunas of Carboniferous age the Productidæ are represented principally, often to the exclusion of all other types, by the great genera Chonetes and Productus. This is much the case in the Guadalupian fauna, though we have in addition to the genera mentioned two others, Strophalosia and Aulosteges. The group of shells separated by Waagen under the title "Marginifera," which usually occurs more or less abundantly in the later faunas of the Carboniferous, appears to be absent from the one under consideration. Several Guadalupian types have much the external appearance of the Marginiferas, but so far as known lack the characterizing internal structures, while one species which shows some suggestion of the characteristic submarginal ridges is more or less aberrant in point of other internal characters and of configuration. It has therefore seemed inadvisable to refer any Guadalupian species to Marginifera.

I have discussed the Chonetes and Producti under special captions, and propose here to pass in review the remaining Guadalupian Productidæ, as well as those which were found in other faunas but are lacking in the Guadalupe Mountains.

In the Salt Range fauna Waagen recognizes, in addition to Chonetes and Productus, the genera Strophalosia, Chonetella, Aulosteges, and Marginifera. Of Strophalosia he distinguishes eight species, as against two in the Guadalupian. While in a general way the Strophalosias of the two faunas are similar to one another, it would appear that they are much less numerous as well as less differentiated in the Guadalupian. No forms corresponding to the group of S. leplayi Geinitz, represented by S. costata, are known in the American fauna.

Aulosteges occurs in the Productus limestone in two distinct types, each represented by a single species, and a rather marked correspondence can be noted between them and the Guadalupian representatives of the genus. Shumard's species Aulosteges guadalupensis corresponds to the Indian A. dalhousii, while the Indian medlicottianus appears to be represented in Texas by A. medlicottianus var. americanus and Aulosteges sp. The latter two, though to all appearances closely related to one another, occur at widely different horizons in the Guadalupe Mountains. Two other Guadalupian types, represented by A. magnicostatus and Aulosteges sp. b, seem not to occur in the Salt Range.

The genus Chonetella, which is represented in the Salt Range by abundant individuals though by only a single species, is not found in the Guadalupe Mountains. Marginifera also probably does not occur there, while Waagen found six species in the Productus limestone fauna. It is true that some Guadalupian species of Productus externally resemble certain of Waagen's Marginiferas, but so far as it has been possible to determine they do not belong to the same genus. Thus the Indian shells assigned to the group of Marginifera splendens may be compared with Productus popei, P. indentatus, and P. texanus; Marginifera transversa, representing the group of M. helica, with Productus latidorsatus; and Marginifera ovalis, of the group of M. spinosicostata, with Productus subhorridus var. rugatulus.

In his paper on fossils from Kashmir and Spiti, in the Himalaya, Diener described Marginifera and Strophalosia, in addition to Chonetes and Productus. Marginifera himalayensis n. sp., the only representative of the genus, has no cognate form in the Guadalupian fauna, though superficially it is of the general expression of certain of the Producti—the smaller ones of the semireticulatus type. Of the two Strophalosias one, Strophalosia aff. S. costata, has no corresponding form in my fauna, but Strophalosia cf. ? tenuispina Waagen is more nearly of the same general type. In his later paper on the Spiti fauna, in which he discriminates between an upper and a lower horizon, this author cites M. himalayensis, above referred to, and Aulosteges gigas from the upper division. The latter appears to resemble A. guadalupensis Shumard.

The Chitichun fauna as represented in Diener's earlier paper, aside from Productus and Chonetes, contains the genera Marginifera and Aulosteges. Marginifera is represented by the Salt Range form M. typica, and is, so far as known at present, absent from the Guadalupian. Aulosteges tibeticus, the single species of Aulosteges cited, is closely related to A. magnicostatus of the Guadalupian. The later paper on Chitichun contains no additional species.

The fauna from Malla Sangcha contains of this group, aside from Productus and Chonetes, only two species of Marginifera—M. typica and M. helica. If these have any corresponding forms in the Guadalupian it must be among species which I have felt constrained to refer to the genus Productus. The same applies to M. himalayensis, cited from the Productus shales of the Lissar Valley, the only species aside from Chonetes and Productus belonging to the Productidæ known from this locality. Marginifera typica has the same relation in the fauna from Byans.

On the whole it can not be said that the Guadalupian fauna, either in its Producti and Chonetes, which have been considered elsewhere, or in the other Productidæ, is very closely related to those of the Salt Range and the Himalaya, though both the resemblances and the differences are interesting. Too much stress should not be laid on the apparent absence of Marginifera from the Guadalupian fauna, for the shells which might be supposed to represent this genus are not always preserved so as to show the distinctive characters. If what appears proves actually to be the case, however, the absence of Marginifera, as of Chonetella, will be among the most important differences, and the most striking resemblance will be found in the genus Aulosteges, which is, however, much better developed in the American fauna.

Kayser distinguishes two species in the Lo Ping fauna, which he refers to Strophalosia, although it is possible that they really belong to Aulosteges. His figures, in fact, may include three species, none of which is closely related to those of the Guadalupian fauna, so far as the latter are known. One of Kayser's species of Productus can probably be placed with Marginifera, namely, Productus nystianus var. lopingensis.

In the fauna from Kantschoufu Loczy cites a little productoid as Productus (Marginifera) longispinus Sowerby, which is probably a small Productus of the semireticulatus group, not very unlike some of the Guadalupian species, though apparently belonging to an older fauna. The forms cited as Chonetes or Daviesiella sp. nov. and Chonetella dubia n. sp. have no related forms in the Guadalupian. The form from the "Permo-Carboniferous" of Batang which Loczy compares with M. ovalis Waagen suggests some of the Guadalupian Producti, e. g., Productus latidorsatus. The form from the Lantsankiang Valley, described as Marginifera desgodinsi, in which the submarginal band is very pronounced, has no Guadalupian representative. The "Permo-Carboniferous" fauna from Tschungtjen, in the province of Yunnan, contains a species which appears to be closely related to Aulosteges medlicottianus var. americanus. Loczy cites it as Aulosteges aff. medlicottianus.

Compared with that of the Productus limestone of the Salt Range, the other Asiatic faunas, even those of the Himalaya, are known in a relatively scanty and imperfect manner. In so far as they involve the minor genera of the Productidæ they show no marked affinity with the Guadalupian fauna; indeed, rather the reverse.

Etheridge senior recorded two species of Strophalosia from Queensland (one of them as a Productus), and Etheridge junior also recognized two from the Bowen River coal field of the same province. The latter are Strophalosia gerardi King and S. clarkei, one of those which the senior Etheridge had described as a Productus, as above noted. The other form mentioned by the latter is cited merely as "Productus" or "Strophalosia." It appears to be of the general type of Aulosteges guadalupensis, though not closely related to it. The imperfectly known S. clarkei is apparently related to A. medlicottianus var. americanus, while the shell referred to S. gerardi seems to have no corresponding form in the Guadalupian, unless perhaps Aulosteges sp. b proves to resemble it.

In the earlier faunas of the Russian section the Productidæ are represented by the genera Productus and Chonetes (and Daviesiella), with an occasional occurrence of Marginifera. In the Gschelian, among the less prominent members of the family Tschernyschew recognizes one species of Aulosteges—A. dalhousii, which is closely related to A. guadalupensis Shumard. Proboscidella claims three species. This group appears to be absent from the Guadalupian fauna, though some of the Guadalupian Producti of the semireticulatus group are like immature stages of the Gschelian species. Eight species are assigned to Marginifera. This group also seems to be missing from the Guadalupian fauna, and although some of the Guadalupian Producti resemble the Russian Marginiferas (chiefly small members of the semireticulati, like Productus popei, P. indentatus, and P. texanus), they are without the characteristic submarginal ridges, so far as known. Marginifera juresanensis looks something like Productus latidorsatus, but here again the Guadalupian species seems to be a true member of Productus. The Guadalupian productoid which seems most likely to prove a Marginifera (Productus? pileolus) is unlike the Gschelian Marginiferas, of which the nearest is unquestionably M. juresanensis.

From the Artinsk Sibirzew cites Strophalosia sp. and Tschernyschew Marginifera typica and Marginifera? spitzbergiana. Krotow found a greater variety in the Artinsk than the other authors consulted, identifying Strophalosia horrescens?, S. morrisiana?, and three species of Chonetella. The latter name was subsequently changed to Chonetina, Waagen having published the same name in the same year. It is perhaps not safe to express an opinion without having had access to specimens, but the generic diagnosis of Chonetella Krotow and the figures of the species leave one very doubtful of the validity of the name even as a subgenus. Strophalosia horrescens? is probably related to S. guadalupensis, but S. morrisiana? is presumably without a Guadalupian representative. In the Permian Strophalosia [Aulosteges?] horrescens and Aulosteges wangenheimi seem to be most often present among the minor productoid genera, and Netschajew describes also Strophalosia fragilis and Aulosteges gigas. Aulosteges gigas and Strophalosia [Aulosteges?] horrescens may be compared with Aulosteges guadalupensis, and Strophalosia fragilis with S. cornelliana and S. hystricula. The types represented by the Guadalupian forms Aulosteges medlicottianus var. americanus, Aulosteges magnicostatus, and Aulosteges sp. b seem not to occur in the Russian Permian. The Russian A. wangenheimi seems to be usually in poor condition, and I am in doubt as to what its specific characters really are.

In the case of these minor genera the Guadalupian fauna appears to be related to the later faunas of the Russian section—perhaps more to the Russian Permian than to earlier horizons. I do not know whether especial importance should be attached to the dying out of the Marginiferas above the Gschelian stage, in which respect a Guadalupian peculiarity is suggested, or whether this should be regarded merely as a phase in the general decline of the brachiopod representation.

From Balia Maaden, in Asia Minor, Enderle cites a form under the title "Strophalosia? aff. horrescens," which he is uncertain whether to call a "Productus" or a "Strophalosia." If it is not a Productus it may be compared with Aulosteges guadalupensis.

Among the Armenian forms which Abich refers to Productus are a number that Waagen and later authors have recognized as belonging to Marginifera. These in many cases closely resemble Guadalupian types which I have referred to Productus (P. latidorsatus, P. subhorridus var. rugatulus, and P. walcottianus). It can not be stated definitely that these Guadalupian species are without the submarginal ridges of Waagen's genus, but I have seen no evidence for believing that they possess them. In the Armenian species these ridges are extremely well developed.

The fauna of the Carnic Fusulina limestone seems to contain, according to Schellwien, only Marginifera pusilla to represent the minor Productidæ. This form is not closely similar to any Guadalupian species, even if some of them should prove to be Marginiferas. In the Trogkofelschichten he distinguishes Marginifera longispina, M. longispina var. lobata, M. pusilla, and M. carniolicus. The figures of Marginifera longispina and the variety naturally resemble certain of the Guadalupian Producti. So also do part of the figures of M. pusilla, the others being different and more like the original figures of the species. In M. carniolicus can be traced a rather distant resemblance to Productus? pileolus, the Guadalupian shell, which at present seems to show most liability of being a Marginifera. Schellwien also figures as ?Aulosteges tibeticus Diener a specimen which is closely related to Aulosteges magnicostatus.

In the Dyas the remarkable diminution in Productus forms is in a measure compensated by the great development of cognate genera. Geinitz recognizes eight species of Strophalosia, most of which may prove to belong in Aulosteges. In addition, the figures of Productus latirostratus are certainly suggestive of Aulosteges. It is probable that several of the Dyas species are of the same general type as A. guadalupensis. Some of Geinitz's figures of Strophalosia morrisiana seem to warrant a comparison with Aulosteges medlicottianus var. americanus and Aulosteges sp. a, although others are very different. Indeed, the appearance of these forms is as various as their manner of preservation. Strophalosia lamellosa seems to belong to the same type as my Aulosteges sp. b, but Productus latirostratus, if, as I suspect, it is an Aulosteges, appears to have no corresponding Guadalupian form.

The Permian fauna of England, closely related to the Dyas of central Europe, is likewise distinguished for its abundance of Strophalosias. King discriminates four species and figures many specimens. The prevailing forms seem to be of the same general type as Aulosteges guadalupensis, but Strophalosia morrisiana may be compared in certain particulars with Aulosteges medlicottianus var. americanus.

In the Arctic faunas of Spitzbergen, Nova Zembla, etc., the Productidæ are in the main restricted to the two genera Productus and Chonetes. It is true that some of these forms may prove to belong to the subsequently established group of Marginifera, but in this case, as in others, the fact can hardly be determined from the evidence at hand. In several instances, however, Toula has cited species of Strophalosia among these forms. One of these is an undetermined species from the south point of Spitzbergen, which is more similar to Aulosteges guadalupensis than to any other Guadalupian type, but at best is only in a general way like it. From the Hornsund he cites Productus (Strophalosia) cancrini and Strophalosia leplayi. The latter, if not a Productus, is probably closely allied to Aulosteges magnicostatus. The other shell seems to be a true Strophalosia, having, however, a sculpture much resembling Productus cancrini, which I take to be a Productus as originally described. Strophalosia cancrini of Toula therefore would seem to be related to S. morrisiana of the Dyas and to be somewhat similar to Aulosteges medlicottianus var. americanus of the Guadalupian.

In the South American faunas the only genus besides Chonetes and Productus (unless among the latter there prove to be representatives of Marginifera), the citation of which appears in the volumes consulted, is Strophalosia, and the only recorded species Strophalosia cornelliana, a form which has two cognate species in the Guadalupian. S. cornelliana was described by Derby from Brazil and he is also the author of several species of Productus which may prove to belong to the subsequently defined genus Marginifera. P. rhomianus is the one most likely to be transferred.

In North America the minor genera of the Productidæ appear only in Aulacorhynchus, Strophalosia, and Marginifera. The rare genus Aulacorhynchus is, so far as known, confined to the Pennsylvanian of the Mississippi Valley and Appalachian region. The only Pennsylvanian species of Strophalosia, S. spondyliformis, is related in a general way to the Guadalupian Strophalosias, but the four Guadalupian species of Aulosteges find no types at all corresponding to them in the Pennsylvanian. Marginifera constitutes an element of greater persistence than variety in the Pennsylvanian faunas, and the representatives of the genus resemble some of the Guadalupian species of Productus in which the characteristic internal features of Marginifera appear to be lacking. The only Guadalupian species which at present seems likely to prove a Marginifera is very unlike any Pennsylvanian member of the genus.

Except for the genus Marginifera, which is scantily developed, if at all, in the Guadalupian but well developed elsewhere, the Guadalupian fauna in its minor productoid genera more closely resembles that of the Salt Range and of the higher Paleozoic terranes of Russia and central Europe than any fauna of the Mississippi Valley.

Genus CHONETES Fischer de Waldheim.

Waagen justly directs attention to the differentiation shown by this genus in the Productus limestone fauna, in which he recognizes as many as 14 species, the development being in some respects no less singular than it is varied. Some estimation of this really remarkable differentiation may be reached by recalling that in the Guadalupian fauna but 4 species have so far been distinguished.

Waagen divided the Salt Range species into three groups, which he distinguishes as the læves, the striati, and the grandicostati. Of these the group of grandicostati, represented by six species, is practically confined to the Himalayan region. The Guadalupian species seem about equally divided between the striati and the læves, for I would place Chonetes subliratus and Chonetes sp. in the former group, while Chonetes permianus and C. hillanus probably belong to the latter. In certain conditions of preservation or over certain portions of the surface there is an ambiguity about these forms which renders them difficult to place, for owing probably to the development of rows of radiating punctæ, they sometimes appear to be obscurely lirated. Thus, while C. permianus usually appears entirely devoid of radiating ornamentation, marginal portions of some specimens show obscure liræ. In this case the appearance may be due to weathering. In C. hillanus, also, obscure liration is developed by exfoliation and sometimes seems to be present toward the margin where unexfoliated. In the latter case the radiating lines may be due not so much to unevenness in the outer surface as to internal structures, such as rows of spinules or of punctæ.

As already remarked, the singular group of the grandicostati is not represented in the Guadalupian and the striati are represented by but two species. The only determinable Guadalupian species of the latter group, C. subliratus, has no counterpart in the Productus limestone fauna. A closer correspondence, however, is shown by the læves, C. ambiensis being represented in the Guadalupian by C. permianus and C. trapezoidalis by C. hillanus, although the latter is more like our common C. geinitzianus than any of the Salt Range species. The singular C. bipartitus has no analogous form in the Guadalupian.

In the Permian fauna from Kumaon and Gurhwal Diener distinguished but two species, both of the striate group. The læves of the Guadalupian are, therefore, so far as known, not represented there, while C. subliratus, of the striati, finds no analogue among Himalayan shells, which are more nearly like Chonetes sp. In the Himalayan region, further, the genus seems to be unrepresented at Chitichun and at Malla Sangcha.

From the Productus shales of the Lissar Valley Diener identifies Chonetes cf. uralica Möller, one of the striati related to Chonetes sp. of the present work. Again from the Productus shales of Byans he identifies C. lissarensis and C. transitionis Krotow, both of which, if they have any representatives in the Guadalupian, find them among shells referred to Chonetes subliratus and Chonetes sp. Diener describes the surface of C. transitionis as consisting of very numerous radiating lines crossed by more distant concentric striæ, remarking, further on, that this combination of radiating and concentric sculpture is so peculiar that a new group name—striati concentrici—must be introduced. Most of the American species which I have supposed should be placed with the striati possess what appears to be this same type of sculpture, though as a rule the crenulating concentric liræ are considerably finer than the radiating ones. This feature is frequently omitted in descriptions, but I had supposed it to be present in all well-preserved representatives of this group.

From the upper Anthracolithic beds at Spiti only C. lissarensis is cited in Diener's second paper dealing with this fauna, but in that on fossils from Kashmir and Spiti he distinguishes four species of Chonetes. Two of these—Chonetes cf. lissarensis and C. austenianus—are referred by him to the striati, and the two others—C. grandicostatus and C. barusiensis—to the grandicostati. Waagen considered C. austenianus also a representative of the grandicostati, a group which, as already remarked, appears to be unrepresented in the Guadalupian fauna. In any event, the only Guadalupian species which are related to them are those here designated Chonetes subliratus and Chonetes sp.

This Kashmir fauna had been partially described by Davidson as early as 1866. He found some of the species afterwards cited by Diener, as well as certain others. The species of Chonetes recognized by Davidson are, among the læves, Chonetes lævis itself, most nearly related to C. permianus Shumard; among the striati, C. hardrensis var. thibetensis, similar in a general way to C. subliratus, and, among the grandicostati, C. austenianus and C. barusiensis. The latter was described as a Spirifer. C. austenianus, as above noted, is placed by Diener among the striati, but it is doubtful if either of these imperfectly known forms has a cognate species in the Guadalupian.

From Niti Pass, in the northern Himalaya, Salter describes only one species of Chonetes—C. vishnu, a coarsely ribbed type with a deep sinus, to which the most closely related Guadalupian forms are C. subliratus and Chonetes sp., though the relationship is rather remote.

In the Carboniferous of Turkestan Romanowsky distinguishes four species of Chonetes—C. hemisphericus Sem., C. variolatus D'Orb., C. kutorganus Sem., and C. glaber Geinitz. The three first-mentioned belong to the striati, and find analogous species in the Guadalupian in C. subliratus and Chonetes sp., though in fact many of the striati, even though from different horizons and different faunas, are in a general way very similar to one another. C. glaber finds a nearly related form in C. permianus. The shell from Turkestan, it may be remarked in passing, is probably distinct from C. giaber = C. geinitzianus, of the Pennsylvanian of America.

Richthofen mentions no species of Chonetes in the Lo Ping fauna of China. In the fauna from the vicinity of Kantschoufu Loczy identifies a number of Chonetes forms, besides a species which he refers to Chonetella and one which he thinks may belong to Daviesiella. The two latter have no analogous Guadalupian species, but the Chonetes representation is more similar. Among the striati he recognizes C. pseudovariolatus Nikitin, C. uralicus var. pygmæus Loczy, C. femingi var. gobicus, and Chonetes cf. buchianus. These are related to C. subliratus and Chonetes sp. of the Guadalupian, C. subliratus being especially comparable to C. uralicus var. pygmæus. The remaining species cited by Loczy, Chonetes cf. politus McCoy, is one of the læves and resembles C. permianus.

The other Chinese faunas described by Loczy are less completely known, and in only one is the genus Chonetes represented—that from Youngtschangfu—in which he cites C. papilionaceus. No species comparable to this very finely lirate form is known from the Guadalupian.

In the upper Carboniferous fauna of Padang, described by Fliegel, the genus Chonetes fails of representation, as also in that from the west coast of Sumatra, described by Roemer. In the Permian of Timor and Rotti, in the Indian Archipelago, Rothpletz cites Chonetella nasuta, a type unknown in the Guadalupian, but no representative of Chonetes. The genus fails to occur also in the small collection from Vladivostok described by Tschernyschew.

From Queensland and New Guinea Etheridge cites five species of Chonetes, only one of which is identified. All are of the striated type, which is imperfectly represented in the Guadalupian by Chonetes subliratus and possibly by Chonetes sp.

Among the Carboniferous faunas of New South Wales described by De Koninck but two species of Chonetes are recognized, both of which appear to occur in the lower beds of the Carboniferous section.

From the younger Carboniferous faunas of Armenia, which Abich and later Arthaber discussed, no representatives of Chonetes are known, although Frech cites C. hardrensis from a younger fauna, which, however, has little interest to our present comparisons. Also in the fauna of Balia Maaden, described by Enderle, this genus appears to be without representation.

In the Russian Carboniferous section Chonetes shows a fairly large and varied representation. From the Productus giganteus zone I have found three species cited, but this fauna does not especially concern us. From the Spirifer mosquensis zone Trautschold cites only Chonetes variolatus of the striati. C. pseudovariolatus (possibly the same species) has also been cited from this horizon. The Chonetes manifest their greatest differentiation, however, in the Gschelian and Artinsk, while in the Permian, like so many of the Brachiopoda, none have yet been found. In the Gschelian Tschernyschew recognizes no less than 12 species. Five of these would probably be assigned to the division læves—C. alatus, C. morahensis, C. trapezoidalis, Chonetes cf. geinitzianus, and C. mesolobus. Of these the second and third were originally described from India and the last two from North America. The identification of C. geinitzianus, of which Tschernyschew himself was in doubt, is almost certainly in error. The shape is unlike that of our American species, and in the figures the surface appears to have delicate radiating striæ. Furthermore, unless Tschernyschew's figure is poor, one can hardly accept the original of the other species as belonging to C. mesolobus. Instead of having a median fold in the ventral valve, bounded on each side by deep furrows, Tschernyschew's figure seems to show only a broad ill-defined sinus. It is true that in some American specimens the lobation is less distinct than in others, although in typical examples it is of course very strong.

Of the læves identified by Tschernyschew, that referred to C. trapezoidalis, more than the other species, seems to have a related form in the Guadalupian, C. hillanus var. inflatus being very similar to it. The remaining seven species, with one exception, belong to the striati, a group, accordingly, which is better represented in the Russian fauna than in the American. Two of the names cited by Tschernyschew—C. granulifer and C. flemingi—are well-known Pennsylvanian species, and the Russian figures, so far as can be told, indicate a much more nearly exact identity than in the case of the two types belonging to the læves just mentioned. In the Guadalupian this group of the striati is represented only by Chonetes subliratus and Chonetes sp. C. uralicus is, perhaps, nearest to C. subliratus, while the remaining Gschelian species are more similar to the imperfectly known Chonetes sp. The last of Tschernyschew's Chonetes—C. timanicus—has a surface, so far as I am aware, unique in the genus, and represents a distinct division coordinate with the striati, the læves, etc. Nothing like it is known from the Guadalupian.

In the Artinsk this genus appears to have flourished and to have been extensively differentiated. Stuckenberg cites C. variolcaris, C. variolatus, C. uralica, C. productoides, C. alatus, C. solidus, C. transitionis, Chonetes cf. capitolinus, C. sinuatus, and C. artiensis. Of these ten species only two are figured, and only two or three are cited from the preceding Gschelian. In the succeeding Kungurstufe only C. variolaris was identified by Stuckenberg. C. alatus appears to be one of the læves, and its peculiar shape renders it almost unique in the genus. No Guadalupian species can be compared with it. C. productoides, however, is one of the normal striati, like C. variolaris, C. variolata, C. uralicus, and others. Krotow also cites a number of Artinskian Chonetes, viz, C. variolaris, C. variolatus, C. uralicus, Chonetes sp., C. solidus, C. capitolinus, Chonetes sp. nov., and C. transitionis, as well as Chonetella minima, Chonetella artiensis, and Chonetella sinuata. Chonetes uralicus, C. solidus, and C. transitionis, which are the only ones figured, are modifications of the common striate division. To this group probably belong also the three species of Chonetella (a name preoccupied by Chonetella Waagen and later changed to Chonetina), for I am of the opinion that Krotow's genus is not valid on the characters mentioned by its author. Tschernyschew cites from the Artinsk only Chonetes variolaris and C. transitionis, the latter remarkable for its very fine liration. On the whole, the Guadalupian Chonetes are comparable to the development of the genus in the Russian section, especially to that of the Gschelian stage. On account of the large representation of the grandicostati in the Salt Range fauna, the resemblance of the latter to the Guadalupian, in point of this genus, is somewhat inferior to that of the Gschelian, where this group is absent and the other singular type represented by C. timanicus is very rare.

I regret that of the fauna of the Fusulina limestone of Palermo, which in the main seems to be so closely related to the Guadalupian, that portion represented by the genera Chonetes, Productus, etc., as described by Gemmellaro, has proved inaccessible to me.

Schellwien found but two species of Chonetes in the fauna of the Trogkofelschichten, one of which he identified as C. strophomenoides and the other as C. sinuosus. C. strophomenoides, as one of the striati, is more closely similar to Chonetes sp. than to any other Guadalupian form, while C. sinuosus is far removed from any known species of that fauna. Indeed, Schellwien includes it in the group of Chonetes mesolobus, a fact of some surprise, since in the Pennsylvanian the range of this type is in the lower part of the Pennsylvanian section. On more careful consideration of C. sinuosus, however, it appears that it is marked with fine radiating ribs, while C. mesolobus is smooth. Thus, though alike in configuration, one species belongs to the striati and the other to the læves.

In his paper on the fauna of the Carnic Fusulina limestone, besides the species discussed in the foregoing remarks (Chonetes sinuosus, there described under the name Chonetes lobatus), Schellwien recognizes also C. papilionaceus var. rarispina, C. granulifer (probably distinct from the American species), C. latisinuatus, and C. obtusus. With the exception of the last, all these species belong to the striati, and, so far as they have Guadalupian representatives, are related to Chonetes sp. and to C. subliratus. C. obtusus seems to be one of the læves, and is of the general type of C. permianus.

Gortani, in the fauna which he recently described from the Carnic Alps, identifies Chonetes variolatus, C. mölleri var. carnicus, and C. strophomenoides. All belong to the striati, a group which has perhaps no typical representatives in the Guadalupian fauna. C. mölleri var. carnicus appears to be related to the species from the Trogkofelschichten which Schellwien describes as C. sinuosus.

The older Carboniferous fauna described by De Koninck from this same general region contains a representation of this genus, but in general it is too unlike the Guadalupian, and of clearly too different a geologic age, to make further comparisons profitable.

In the Dyas of central Europe, as well as in the closely related Permian of England, the genus Chonetes seems not to occur, a peculiarity which these faunas share with the Permian of Russia.

In his papers on the Carboniferous of Spitzbergen Toula cites a number of species of Chonetes, such as C. verneuilianus var. spitzbergianus, C. granulifer, C. hardrensis, and C. papilionaceus. These all belong to the striati and are more or less related to the Guadalupian species of that group, though it is probable in this case, as in others, that the Guadalupian species C. subliratus would be found to differ from those which have a similar configuration, in the imperfect and faint development of the liræ. Toula also described from Spitzbergen Chonetes capitolinus, one of the læves, a fine large species with slightly developed fold and sinus. C. permianus seems to be the most closely related Guadalupian form. From Nova Zembla also this author cites two Carboniferous species of Chonetes—C. variolatus and C. rotundatus. The former is one of the striati, and is perhaps related to Chonetes sp. of this report. The other belongs to the læves, and is of the general type of C. permianus.

Stache, in describing several imperfectly known faunas from the West Sahara, representing apparently a number of different periods in the Carboniferous, cites Chonetes aff. tuberculatus McCoy from one of them. This species, though it is one of the striati, has no very close Guadalupian allies, and it probably belongs in an older fauna.

Salter did not identify this genus from Bolivia, but Toula obtained three species, which he identified its C. tuberculatus McCoy, C. mucronatus Meek and Hayden, and C. glaber Geinitz (= C. geinitzianus). C. tuberculatus and C. mucronatus, the latter generally regarded as a synonym of C. granulifer Owen, belong to the striati, while C. geinitzianus is one of the læves. The related species in the Guadalupian fauna have already been indicated. D'Orbigny also, in his material from Bolivia, described a species belonging to this genus—the much-cited C. variolatus. It is one of the striati, and if it has a related form in the Guadalupian fauna it is Chonetes sp.

In his work on the Brazilian faunas Derby distinguishes two species of Chonetes, both belonging to the læves. One he calls C. amazonicus and the other C. glaber. It is doubtful if the shell identified as C. glaber (= C. geinitzianus) is really the same as the North American species; it appears to be closely related to C. amazonicus. These forms seem to be in a measure intermediate between C. permianus and C. hillanus.

The only other South American country from which the Carboniferous is known, so far as I am aware, is Peru, and in this imperfectly known fauna Gabb recognized no species belonging to the present genus.

It is uncertain how many species of Chonetes should be recognized in the upper Carboniferous of North America. In his valuable bibliography Weller lists nine, although other names have been introduced which he relegates to synonymy. Most of these species are modifications of the striate type, which is much less well developed in the Guadalupian. While the striati were persistent from the Devonian until the genus ceased to exist, the læves I have come to regard as a subsequent development and as conditionally indicating rather late Carboniferous time. At least such seems to be the case in the American sequence, so far as known. Perhaps an exception should be noted in the case of C. mesolobus, which I think we must regard as a member of this group, but the configuration of this species is so peculiar as to render it almost sui generis, the foreign identifications being, so far as I have surveyed them, very questionable. Some of these striate Chonetes of the Pennsylvanian naturally resemble the Guadalupian forms, just as similar types can be picked out in very many faunas. The læves comprise among the Pennsylvanian Chonetes, aside from Chonetes mesolobus, only C. geinitzianus, whose most closely related Guadalupian species is C. hillanus. The western form described by White as C. platynotusis also probably one of the læves, although it shows traces of radiating lines. In configuration and surface it is distinct from either C. geinitzianus or C. mesolobus and more nearly allied to the Guadalupian group, especially to C. permianus.

CHONETES PERMIANUS Shumard.

Pl. XX, figs. 1 to 3a; Pl. XXIX, figs. 1 and 2.

1859. Chonetes Permiana. Shumard, Trans. Acad. Sci. St. Louis. vol. 1, p. 390 (date of volume, 1860).

[Permian]: Conglomerate at mouth of Delaware Creek, Texas.

Shell small, subsemicircular, widest at the cardinal border, width one-third greater than the length, front and sides rounded. Ventral (receiving) valve moderately convex, without mesial sinus; cardinal margin sloping gently from beak to extremities and marked with five or six spines; ears mucronate, gently convex, and separated from the vault by a gentle depression. Ventral valve and area unknown. Surface marked with extremely fine concentric striæ of growth.

I have several specimens of this species before me, none of which exhibit any traces of longitudinal striæ.

Found in the conglomerate at the mouth of Delaware Creek, Texas.

The foregoing description, which is taken complete from Shumard, conveys an adequate idea of the characters of this shell, and I am able to add little to it and to change but little. It seems nearly certain, however, that Shumard inadvertently used "ventral" for "dorsal" in the sentence "Ventral valve and area unknown." His statement that the ears are mucronate is difficult to understand, if my fossils rightly belong to his species, for the cardinal angle in them seems to be regularly a right angle, seldom extended, and never produced into mucronate points. The largest specimen in my collection measures 14 mm. across, several measure 11 mm., while the average is somewhat smaller. I find that in some examples the width is a little more than one and one-third times the length. The other characters mentioned agree so closely with my specimens, which differ from the other Chonetes found at the same general horizon in just those particulars—i. e., the subcircular shape, absence of sinus, and absence of radiating liræ,a together with the presence of faint concentric ones—that the identification certainly appears to be correct.

---

aIn some large specimens near the margin, especially on the under layers of the shell, the presence of liræ can be distinctly made out.

---

Shumard's specimens came from pebbles in the conglomerate at the mouth of Delaware Creek, the source of which probably was the limestones of the Guadalupian. In my collections from the Guadalupian section this form occurs only in Shumard's "dark limestone." It has been subsequently found by Mr. Richardson at several points considerably south of the south end of the Guadalupe Mountains, a fact which probably has some significance with reference to correlation. C. permianus is evidently related to C. ambiensis Waagen, though probably not identical with it. At all events Shumard's name long antedates that used for the India species.

Chonetes permianus is very suggestive of a species which was later described by White as C. platynotus. That form, however, shows a slight sinus and possesses radial markings that are either very faint liræ or the effect produced by the linear arrangement of the interliral pores.

Another American species which is related to this is C. geinitzianus Waagen.b A manuscript footnote in Meek's copy of Geinitz's "Carbon formation und Dyas in Nebraska" suggests a comparison of that species with C. permianus. The resemblance is certainly marked, but does not extend to specific identity, for the Nebraska shell is more transverse and more mucronate; there is always a well-marked sinus, and the cardinal spines are more numerous, since Meek mentions 8 or 10 to 12 or 14, while Shumard records but 5 or 6 in C. permianus, apparently for the whole area, an observation which is in accord with my own specimens. These seem to indicate also that the concentric growth lines are more apparent in C. permianus, in which they are pronounced and strongly lamellose near the margins, with the small punctæ mentioned by Meek as occurring in C. geinitzianus either absent or restricted to the anterior portions of the shell; but these characters are so minute as to be greatly affected by preservation, and I am unwilling to maintain this difference without examining more material under different conditions of preservation. However, I feel little doubt that C. permianus is distinct from C. geinitzianus.

---

bIt seems clear that this name should be employed for the species described by Geinitz as C. glaber, since the latter name is preoccupied by C. glaber Hall.

---

Horizon and locality.—"Dark limestone," Pine Spring (station 2930), and east of Guadalupe Point (station 3762b), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2936 and 2969).

CHONETES HILLANUS n. sp.

Pl. XI, figs. 8 to 10.

1859. Chonetes Flemingi (?). Shumard (non Norwood and Pratten), Trans. Acad. Sci. St. Louis, vol. 1, p. 390 (date of volume, 1860).

White [Permiam] limestone: Guadalupe Mountains.

Shell rather large, very transverse, subtriangular. Ventral valve somewhat strongly convex. Sinus well marked; hinge line much produced, alate. Wing depressed, defined by a broad, obscure sulcus, which is stronger on the inner but indistinct on the outer side. Area rather high. Number of spines not known. In young specimens the convexity is less strong, the shell more transverse, and the sinus less deep.

Dorsal valve known only as an external mold. It is transverse, strongly concave, and in other respects resembles the ventral.

Surface marked by very fine indistinct liræ, which are broad and flat, their tops flush with the general curvature of the shell, defined by fine obscure striæ.

It is probably this species which Shumard cites as Chonetes flemingi?, remarking:a

The fossil from the white limestone of the Guadalupe Mountains corresponds pretty well with the figures and description of the above-cited species, though I can see minor points of difference which leave me in doubt as to whether it is really identical.

---

aTrans. Acad. Sci. St. Louis, vol. 1, 1858-1860, p. 390.

---

It seems to me that there can be no doubt as to the distinctness of this shell from C. flemingi. It is more transverse and alate, with a stronger fold and sinus. The ventral valve is more convex. The liræ are fainter and appear to lack the tubulose-spinose character of C. flemingi.

In my original list of this fauna a certain resemblance between this form and C. trapezoidalis Waagen is suggested. The resemblance, however, was more in the shape and general appearance than in detail. A close comparison is hardly necessary, but C. hillanus has liræ, though very faint ones, shallow sinus, and no lateral plications.

I take pleasure in naming this species for my friend Mr. B. F. Hill, who aided me in the collection of this fauna.

This species is best represented in the Capitan formation of the Guadalupian (station 2926), where it is fairly abundant. Two specimens possibly belonging to it were found near the base of the series, in Shumard's "dark limestone," associated with Chonetes subliratus.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

CHONETES SUBLIRATUS n. sp.

Pl. XX, figs. 4 to 7.

Shell of medium size, transverse, alate, tumid. Sinus strong. The shell rises, as it were, into two plications, one on either side of the sinus, falling with a rapid descent to the wings, which are depressed and convex. Number of spines not ascertained.

Surface marked by moderately fine ribs, which are depressed and indistinct, crossed by fine, rather strong, somewhat irregular concentric liræ. The sculpture in brief recalls that of C. ornatus of the Kinderhook, but it is on a much smaller scale.

As compared with C. hillanus, the greater convexity, more subquadrate shape, and stronger sinus make the general appearance of well-characterized specimens very different. The ribs are somewhat more distinct, more numerous, and not so flattened, and the concentric liration is stronger.

In young specimens the sinus is less strong and the convexity more moderate. The chief difference in configuration between them and similar stages of C. hillanus is that they are more quadrate and less transverse. In the very young example represented by fig. 7 the sinus is practically absent, yet the convexity is considerable.

The typical examples of this species were obtained from Shumard's "dark limestone" (station 2930), where they are associated with C. hillanus. A single specimen from nearly the same horizon was obtained at station 2906, southwest of Guadalupe Peak, and another from the Delaware Mountain formation, at station 2919. The identity of the last two examples is somewhat doubtful. That from station 2906 has the alate, subquadrate, inflated valves and general configuration of the typical examples, but with the sinus faint or reduced to a mere flattening. The terminal spine at the end of the hinge is very long, slightly curved, and directed to the cardinal line at an angle of about 1350. The configuration of the specimen from the Delaware Mountain formation, which is that of the typical examples, though less strongly marked, may be the result of compression. The real affinities of this shell are possibly with Chonetes sp.

There seems to be a tendency in these shells toward an evanescence of the sinus and a lowering of the general convexity. Therefore some of them simulate Chonetes hillanus in general configuration, but the sculpture should serve to distinguish them if it is not obscured. The Guadalupe specimens are apt to be exfoliated, and I have referred to C. hillanus several examples from the "dark limestone" which may really belong here. One of them especially is very similar to the original of fig. 3 on Pl. XX, but is less convex and with a fainter sinus. In C. hillanus the ribs are relatively broad and flat, separated by obscure striæ or possibly only by rows of pores. In C. subliratus the ribs are also faint, but they are thin, with relatively wide striæ between, and the concentric liræ are stronger than in C. hillanus.

Horizon and locality.—Base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930); Delaware Mountain formation, Guadalupe Point (station 2919), Guadalupe Mountains, Texas.

CHONETES sp.

In the yellow sandstone of the Guadalupe section at station 2931 were obtained a few specimens of Chonetes whose preservation as internal casts makes it impossible to ascertain their relation to other species. Apparently they can not be referred to the other species recognized in this report. The width of one specimen was 18 mm. and the length 11 mm. Of another, more fragmentary example, the width must have been 22 mm. and the length 13 mm. The shape was semicircular, the hinge line as long as or a little longer than the shell in front. The convexity was slight and regular, there being no sinus. The cardinal spines appear to have been rather numerous, probably 6 to 8 on each side. The surface is unknown. The rows of internal spinules may or may not indicate a striated surface. If striated, the liræ were probably rather fine.

A specimen from about the same horizon, from station 2903, has also been referred to this species. It is slightly more convex, but otherwise the configuration is similar, and it is ornamented with distinct though fine ribs. It is possible that another specimen from station 2919 which I have referred to C. subliratus may also belong to this species, the greater convexity and different configuration being the result of lateral compression which it has evidently undergone. I have referred here also a fragment from the black limestone at the base of the Guadalupe section. So little is known about either of the forms involved that their specific relationship to one another can not be ascertained.

The low convexity of this shell, comparatively short hinge, and absence of fold and sinus distinguish it from typical C. hillanus or C. subliratus. For all that is known it might belong to C. permianus, but the much greater size is suggestive that if the exterior were preserved other distinctive characters would be found. Though hardly so mucronate, this species is suggestive of young examples of C. granulifer, but besides being smaller and lacking cardinal extensions, it has a much larger septum than Owen's species.

Of course if the striated specimen from station 2919 really belongs to this species the presence of ribs would at once distinguish it from C. permianus and to a less degree from C. hillanus and C. subliratus.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point (stations 2903 and 2931); basal black limestone, Guadalupe Point (station 2920), Guadalupe Mountains, Texas.

Genus PRODUCTUS Sowerby.

The faunas of the Carboniferous are characterized by no other type so highly as by the genus Productus. These shells are usually present in abundance, and they manifest the greatest diversity of sculpture and configuration. Considered as a whole the group has shown unusual placticity, developing not only widely different types, all referable to the same genus, but also abundant intermediate stages between what one would suppose to be wholly distinct species. In consequence, specific discrimination among the Producti has always been a difficult matter, and authors have shown wide differences of opinion as to where the limits of species should be drawn. In spite of a frequent tendency to group really unlike forms under a single specific title because of transitional varieties, a large number of specific types have been recognized and named. In his monograph on the genus Productus,a De Koninck in 1847 listed 64 species, based to a considerable extent on European types,b and he did not cover all the ground. Probably at the present time this number would have to be trebled.

---

aRecherches sur les animaux fossiles, pt. 1, Liége, 1847.

bWhether first described from European areas or not, only 5 of the 64 classified species were cited as non-European.

---

Because of the great number of mutations which the group affords and the uncertainty as to what in most cases should really be included in the limits of a single species, a comparison between the Guadalupian fauna and those of other areas will best be made in terms of more or less distinct groups which have been recognized within the genus.

In the work above referred to De Koninck proposed a classification of the Producti, an effort rendered necessary even at that time by the importance of the genus and its copious specific representation, and he was so far successful that his scheme, more or less modified, is still in use. It is perhaps better regarded as a key than as a classification, and appears to possess the imperfections usual to keys if considered as classifications. Species subsequently described from epochs or areas whose faunas were unknown or little known when his monograph was written do not in all cases fall readily into it. Several instances have been met within the present fauna, as one of which I would regard those strongly ribbed shells which are closely related to the semireticulati, but whose posterior portion is more or less completely without concentric wrinkles.a Another comprises shells which have a posterior portion crossed by strong concentric wrinkles, but are without ribs.b

---

aThese would, I suppose, go with Productus cera and P. giganteus among the striati, but the association would hardly be a natural one.

bDe Koninck himself pots a species of this type (P. sublævis) with the semireticulati, but this is not in conformity with the wording of his classification.

---

Another defect in De Koninck's classification, though this is less demonstrable and more a matter of personal opinion, is that it is in some respects at least artificial, assembling unrelated forms and separating related ones. Thus our western species Productus multistriatus and the Russian form P. mammatus, which resembles it, fall into De Koninck's division striati, but to me they certainly do not appear to be related to P. cora. On the other hand, Productus pertenuis and certain other species which really do seem to be related to P. cora would have to be assigned to the spinosi.

Finally, the divisions appear to be of very unequal importance. The caperati and horridi are distinguished merely by the presence or absence of a sinus, a feature which often varies greatly between individuals belonging to the same species. My point is made if the relation between these two groups is compared with that, say, between the semireticulati and the striati.

It is far from my thought to invent a new classification for the Producti, but were I to do so I would endeavor to establish it on certain family, or perhaps I should say gentile, differences, which are sometimes well marked in this group. In some instances these relations have been observed in De Koninck's classification and make the permanent part of it. Possibly they form the basal idea of a whole, an idea which has been obscured by a too brief characterization of some of the groups, and if one adhered more to the spirit and less to the letter of the matter the faults which I have mentioned, with the exception of that of incompleteness, would be remedied. Thus the caperati of De Koninck chiefly comprise Devonian and early Carboniferous shells for which Hall established the genus Productella, and they are quite distinct from the Permian Productus horridus, which probably gave name and conception to the horridi, but De Koninck's fundamentum division is, or his statement of it, is such that some of the Productellas, and some of the Marginiferas, too, would be grouped with P. horridus.

Nevertheless, a classification of some sort is a great convenience, and De Koninck's, in spite of what seem to me imperfections, is the best available, so I have employed it here, with certain modifications introduced by Waagen, following in a few cases, however, what I take to be the leading of the natural relations rather than the literal description of his divisions.

In the Guadalupian fauna 25 varieties of Productus have been recognized, which may be distributed somewhat as follows:

LINEATI.

Group of Productus neffedievi.

SEMIRETICULTI.

Group of Productus semireticulatus.

Group of Productus popei.

Group of Productus guadalupensis.

Group of Productus occidentalis.

SPINOSI.

Group of Productus cancrini.

HORRIDI.

Group of Productus latidorsatus.

Group of Productus walcotticnus.

CAPERATI.

Group of Productus? pileolus.

IRREGULARES.

Group of Productus striatus.

PRODUCTI OF UNDETERMINED POSITION.

The largest number of species belong to the semireticulati, but I have referred to that division types in which the wrinkling of the posterior portion is so indistinct that I suspect De Koninck would have placed them with the striati. Their relationship with P. semireticulatus certainly appears much more close and essential than with P. cora or P. waagenianus. To the group of P. semireticulatus itself but three species have been referred, so that even with the somewhat broad limits with which that group is here interpreted the development of the typical semireticulati is rather slight compared with that in other faunas. Productus semireticulatus var. capitanensis is a typical example of this group, to which the imperfectly known Productus sp. c also probably belongs, as well as Productus mexicanus Shumard, although the exact form which Shumard had in hand is still a matter of doubt. Productus semireticulatus var. capitanensis is the only large Productus known to me from the Guadalupian, the representatives of the genus in this fauna being rather small and somewhat restricted in differentiation.

P. popei stands as the representative of another division of the semireticulati to which I have referred the largest number of species. In these shells the ribs are for the most part well defined, except over the posterior portion. The concentric wrinkles are obscure, so that the visceral area is but faintly marked by any sort of sculpture. These species are, furthermore, distinguished by their small size, high convexity, and generally deep sinus.

P. guadalupensis represents another type believed to be of the semireticulati, though the evanescence of the concentric wrinkles over the visceral region might be regarded as debarring it. This species resembles the group of P. popei, but is much larger.

A fourth type of the semireticulati is represented by P. occidentalis, distinguished by a general evanescence of the characteristic features. Over the anterior portion especially the ribs become large and obscure.

To the horridi have been referred in all five Guadalupian species, which may be divided into two groups, one typified by P. latidorsatus and the other by P. walcottianus. To the former group belong P. latidorsatus, P. latidorsatus var., and possibly the imperfectly known Productus sp. e, while to the latter may be referred Productus subhorridus var. rugatulus, and Productus walcottianus itself. As these forms possess a sinus, more or less distinct, they could not by definition be placed with the caperati, and indeed the impropriety of assigning them to a group corresponding especially to Hall's Productella is rather obvious. But they certainly are not closely related to P. horridus even in the most favorable instance (P. latidorsatus), while the least favorable (P. walcottianus) is very different indeed. They are not typical representatives of the horridi, because of the tendency of the spine bases to pass into continuous ribs and the presence of more or less distinct concentric wrinkles over the posterior portion. It might perhaps be urged that these five species do not make a homogeneous assemblage, and this appears true if only the extremes are regarded, but the intermediate forms are so linked one with another as to require careful discrimination. This row of forms seems in fact to occupy intermediate ground between the semireticulati and the horridi, one extreme, P. walcottianus, being related to the former and the other, P. latidorsatus, related to the latter.

One species has been referred to the caperati, but in this instance also the reference is doubtful. Productus pileolus could be placed in the genus Marginifera with greater propriety than any Guadalupian species known, but it has not seemed to me that the facts warranted doing so. On the other hand, though it must needs be referred to either the caperati or the horridi, from the latter it is debarred by its obvious lack of a sinus, while I feel very doubtful about placing it in a group made up for the most part of Productellas.

Belonging to the spinosi and apparently closely related to Productus cancrini, the Guadalupian fauna furnishes P. meekanus, P. signatus, and P. signatus var. Here again the assignment of these species is attended with difficulty. For my own part I imagine them to be a development of the cora group, but to place them with the lineati is, under the definition, quite impossible. P. meekanus especially, and P. signatus, which is closely related to it, are more or less completely covered with concentric wrinkles. This would seem to warrant referring them to the undati, but the undati are not supposed to be covered with spines, which are a feature in the present forms still more marked than the wrinkles. It is this character, together with their resemblance to certain of the recognized spinosi, such as P. cancrini, that has led me to place these species in that division, although the spinosi are not characteristically wrinkled to any great extent.

Representing the striati we have in the Guadalupian P. waagenianus and P. waagenianus var., and belonging to Waagen's group of the irregulares only P. pinniformis.

The disposition of two species affords more than the usual difficulty. One of these is Productus sp. d, which but for the reflexed margin might belong to the group of P. popei, among the semireticulati, and the other is the singular and imperfectly known Productus limbatus, whose relations are possibly with the same group.

Though the genus Productus is represented in the Guadalupian fauna by a considerable number of species, its development is conspicuously inferior to that of the Carboniferous fauna of the Salt Range of India, where it is so abundant, so large, and so diverse as to have given name to the whole Productus limestone. Though smaller and less varied, the Guadalupian Producti are of interest no less for what is present than for the types which are lacking.

Of the lineati Waagen represents two groups among the Salt Range species, one the group of Productus neffedievi (represented by P. lineatus) and the other the group of P. corrugatus (represented by P. cora). In the Guadalupian P. cora has no near representative. P. waagenianus may be the Guadalupian representative of P. lineatus, but its small size and the very distinct character of the dorsal valve which seems to belong to it renders the relationship a remote one. It is possible that P. guadalupensis should be placed with the lineati, in which case it would probably belong in the group of P. neffedievi, along with P. lineatus.

Waagen recognizes seven Salt Range species among the semireticulati, mostly allied to Productus semireticulatus and P. costatus. This group of large shells is represented in the Guadalupian by only two species (P. semireticulatus var. capitanensis and Productus sp. c), while the other groups of the semireticulati, that of P. guadalupensis, of P. occidentalis, and of P. popei, find no closely related forms in the Salt Range fauna. This is perhaps less true, however, of the group of P. occidentalis than of that of P. popei, since some of the forms included by Waagen under P. gratiosus are very suggestive of some of those included in the group of P. popei, though they have a considerably more distinctly reticulated visceral area. They may prove to be closely allied to the imperfectly known P. mexicanus of Shumard.

The spinosi are represented in the Guadalupian by three species related to P. cancrini, and in the Salt Range by only P. asperulus. In neither case are the forms found in one related to those found in the other fauna.

Of the fimbriati Waagen recognizes six species, and perhaps no greater difference between the Producti of the two faunas can be pointed out than in this instance, for this group, so far as known, is not represented in the Guadalupian at all.

Waagen distinguishes two groups among the Salt Range representatives of the horridi. P. opuntia, of the group of P. geinitzianus, finds its Guadalupian representative in P. subhorridus var. latidorsatus. Neither the Guadalupian group of P. walcottianus nor the Indian one of P. kiangsiensis is quite characteristic of the horridi, and they are also rather unlike one another.

Among the irregulares the Guadalupian and Salt Range faunas have related species in P. pinniformis on the one hand and P. compressus and P. mytiloides on the other.

The two Guadalupian types whose affinities have not been satisfactorily settled, P. limbatus and Productus sp. e, seem to have no allies in the Salt Range.

On the whole, therefore, while the two faunas have some rather striking points of resemblance in the Productus representation, the points of difference are more numerous and more important.

The faunas of the Himalaya seem in a general way to be related to those of the Salt Range, but certain features brought out in Diener's papers may be given comment. These will chiefly have to do with types not found by Waagen.

In the paper on the Permian fossils from Kumaon and Gurhwal Diener finds five species of Productus, three belonging to the fimbriati, a group which, as already pointed out, though well represented in the Salt Range, is absent from the Guadalupian faunas, and two species belonging to the spinosi. The latter represent the group of Productus cancrini, a type which has Guadalupian representatives also and which is absent in the Salt Range, although the section of the spinosi occurs there.

In his first paper on the Carboniferous fossils from Kashmir and Spiti Diener distinguishes no less than 11 species of Productus, which he distributes among eight divisions of the genus—the lineati, the undati, the semireticulati, the spinosi, the fimbriati, the caperati, and the irregulares. The form which he refers to P. cora, among the lineati, is suggestive of that which I have described as P. waagenianus, and probably is wrongly identified with P. cora. In this connection attention may be called to a shell which he figures as Strophomena analoga. This type, of which Leptæna rhomboidalis is the most familiar representative, does not, so far as I am aware, range above the lower portion of the Carboniferous; yet the shell from Kashmir was found associated with Productus abichi, Marginifera himalayensis, and Chonetes grandicostatus, representing a fauna which Diener is inclined to call Permian. This shell so much resembles the dorsal valve of P. waagenianus that it seems permissible to suggest that it may represent this portion of the form referred by Diener to P. cora, the dorsal valve of which was not thought to occur in Diener's collection.

The division of the undati seems to be absent from the Guadalupian fauna, and the form referred by Diener to P. undatus has no Guadalupian representative. As belonging to the semireticulati Diener cites only P. semireticulatus and Productus cf. longispinus. The former has a representative in the fauna under discussion in P. semireticulatus var. capitanensis, and Productus sp. c, and the shell referred to Productus cf. longispinus seems to be related to P. texanus and possibly P. mexicanus. The other groups of the semireticulati found in the Guadalupian are apparently absent from Diener's fauna.

Two species are placed by Diener among the spinosi. Productus cf. scabriculus seems to have no representative in the Guadalupian fauna, while Productus cf. spinulosus may be represented, somewhat remotely, it is true, by P. latidorsatus, though I have placed the latter species among the horridi. The American shells belonging to the group of P. cancrini, among the spinosi, seem to be unrepresented in the fauna described by Diener.

The fimbriati, of which Diener cites three species, are, as already remarked, absent in the Guadalupe Mountains.

I have found only one Guadalupian species which seemed referable to the caperati, to which Diener refers the shell identified as Productus aculeatus. These two forms do not appear to be related to one another to any very marked extent, but perhaps the Guadalupian shells which I have called P. walcottianus and P. subhorridus var. rugatulus are really more closely allied than would appear from their having been placed in another division.

Lastly, among the irregulares the species described by Diener as Productus mongolicus is clearly allied to P. pinniformis. In fact, the irregulares are as a rule much more nearly related to one another than are the members of the other groups. The horridi, to which I have referred certain Guadalupian species, have not been recognized in this fauna, but in some cases they probably find related forms in species which Diener has placed in other divisions.

These fossils seem to have been derived more from Kashmir than from Spiti. Davidson had long before described a suite of fossils from Kashmir in which he recognized nine species of Productus, viz, Productus semireticulatus, P. cora, P. scabriculus, P. humboldti, P. longispinus?, P. striatus?, P. spinulosus?, P. lævis n. sp., and Productus sp. Davidson allowed greater latitude to specific limits than is in my estimation justified, and as some of his identifications are not illustrated it is impracticable to compare such instances with the fauna of the Guadalupe Mountains. It is safe to say, however, that P. scabriculus, P. humboldti, and probably P. spinulosus have no analogues in the Guadalupian, though some of the forms which I have placed with the horridi are perhaps distantly related. The little shell which Davidson describes as P. lævis seems to resemble P. latidorsatus.

In a later paper Diener discusses the fauna of Spiti, dividing it into two stratigraphic sections. The lower fauna contains P. lineatus (perhaps represented in the Guadalupian by P. waagenianus, possibly even by P. guadalupensis), Productus sp. indet., of the group of semireticulatus (also represented in the Guadalupian), Productus undatus (seemingly without any related form in my fauna), P. scabriculus (also without a Guadalupian representative), and P. nystianus var. lopingensis. The latter species resembles the form from the Delaware Mountain formation which I have called Productus sp. e, but the Guadalupian form is too imperfectly known to admit of a safe comparison. To some extent Productus subhorridus var. rugatulus is a related form, and to a considerably less extent P. latidorsatus. From the upper fauna at Spiti Diener cites only one species of Productus—P. gangeticus, a type as yet unknown in the Guadalupian fauna.

Diener twice treats of the Carboniferous faunas of Chitichun. In the first paper are listed nine species of Productus, which he assigns to the lineati, semireticulati, spinosi, fimbriati, and irregulares. The types which he refers to P. lineatus and to P. cora appear from his figures to be the same species, which I can not but believe to be distinct from P. cora, at least the form which in North America it is customary to identify with that species, though truly it is not altogether safe to trust figures in such comparisons.

The semireticulati of the Chitichun fauna, like those of the Salt Range, show in the main the same large species related to P. semireticulatus and P. costatus, of which P. semireticulatus var. capitanensis and Productus sp. c are the only Guadalupian representatives. In a series of specimens referred to P. gratiosus, however, a species which is represented in the Salt Range by a number of mutations, we have some forms closely related to P. popei and its allies. The Indian forms for the most part are larger, with less distinct sinus, and by reason of stronger corrugations over the visceral region more clearly deserve to be assigned to the semireticulati. They are perhaps closely allied to the imperfectly known P. mexicanus, but one of the specimens figured by Diener, which is small, highly arched, and with a deep sinus, might almost have been drawn from a specimen of P. popei. As in the Guadalupe Mountains, so at Chitichun, the spinosi are represented only by species of the group of P. cancrini. In the Salt Range, it will be remembered, the group of P. cancrini is not found, and the spinosi are represented by a quite different type.

The fimbriati, which are not known in the Guadalupian fauna, are represented at Chitichun only by the Salt Range species P. abichi.

The Productus faunas of the Guadalupe Mountains and Chitichun show a community in the representation of their Producti irregulares in the related species P. pinniformis on the one hand and P. mongolicus on the other. With this terminates my comparison of the Guadalupian with the Chitichun Producti, for only the same species appeared among the material on which Diener's second paper was based, and he does not describe them again.

The same author distinguishes six species in the Permian fauna of Malla Sangcha. Productus abichi of course has no Guadalupian equivalent, and, on the other hand, many Guadalupian types seem to be unrepresented in the Himalayan fauna. The large coarsely ribbed shells belonging to the semireticulati which form so noticeable a feature of the Productus limestone fauna, and are represented in the Guadalupe Mountains by P. semireticulatus var. capitanensis and Productus sp. e, seem to be missing in the collections from Malla Sangcha, their place being taken by P. chitichunensis, a species which has no very close relative in the Guadalupian. Of the latter fauna the group of P. popei, also belonging to the semireticulati, seems to be represented by P. gratiosus, a Salt Range species which Diener says is the commonest in his collection and shows great variability. Some of the forms appear to be very similar in configuration to the American ones, but I have not been able to ascertain whether they have the strong corrugations over the visceral region, the absence of which seems to be a peculiarity of our group of shells. Shumard's P. mexicanus is also more or less closely allied to P. gratiosus. The remaining species cited from this fauna are P. mongolicus, one of the cora group, P. undatus, without Guadalupian representatives, and P. planohemispherium, a small, somewhat nondescript species without, so far as known, any corresponding form in the Guadalupian, unless possibly it be P. waagenianus,which is more or less similar, though I doubt if Netschajew's species belongs to the lineati.

From the Lissar Valley Diener cites only two species of Productus, both belonging to the fimbriati, a group not found in the Guadalupian fauna. One of these again is cited from the Productus shales of Byans, the only other Productus obtained being P. cancriniformis, whose Guadalupian representatives are P. meekanus and P. signatus.

In describing the fauna from Niti Pass Salter distinguished but two species of Productus—P. purdoni, one of the fimbriati, and a little shell which he calls P. flemingi Sowerby?. The figure is not very good, but appears to represent one of the spinosi, of the cancrini group, similar to P. meekanus.

The faunas of the Salt Range and of the Himalaya, being considered together, in view of their representation of the genus Productus show many points of resemblance and some of marked difference when compared with the Guadalupian. The most striking differences seem to be the great abundance and variety in the Indian faunas of large shells related to P. semireticulatus and P. costatus, which are represented in a much inferior manner in the Guadalupian, and the presence of numerous species belonging to the fimbriati, a group, so far as known, which is lacking in the Guadalupian fauna altogether.

Romanowsky distinguished 11 species of Productus in his material from Turkestan, some of them of rather unusual type. Productus striatus has an analogous species in the Guadalupian in P. pinniformis. It is somewhat doubtful if the shell identified as P. cora really belongs with that species, and the most closely related form in the Guadalupian is P. waagenianus. The specimen which Romanowsky figures under the title of P. giganteus does not appear to me really to be Martin's species. If it is, its horizon must needs be considerably older than the Guadalupian and older than that of some of the other species of the same report. There is no corresponding form in the American fauna. Of the semireticulati he discriminates four species—P. semireticulatus, P. deruptus Rom., P. boliviensis, and probably P. reticulatus Rom. It should be remarked that the latter name had been long preoccupied by Gabb for a shell from Peru. These semireticulati, especially the first three, find analogous species in the Guadalupian in P. semireticulatus var. capitanensis and Productus sp. c, but the other Guadalupian members of the semireticulati seem to be unrepresented in Romanowsky's fauna. He identifies two of the fimbriati (P. punctatus and P. fimbriatus), however, a group which, as already pointed out, appears to have no representatives in the Guadalupe Mountains. Romanowsky figures a species which he identifies as P. spinulosus and which accordingly should be one of the spinosi, but as it is represented without ribs, though with concentric wrinkles and spines, it is also probably one of the fimbriati and without any closely related Guadalupian form. The last species, described, as P. vlangalii, with its heavy radiating costæ and spines, presents a singular appearance, but can probably be placed in the spinosi. This species also has no analogous form in the Guadalupian, and as the latter contains types not found in Romanowsky's collection, e. g., the group of P. cancrini, it seems to be only distantly related to the faunas of Turkestan, much less nearly than to those of the Salt Range and of the Himalaya.

Thirteen species of Productus were distinguished by Kayser in the Lo Ping collections. The semireticulati are represented by a variety of forms, some of which appear to be closely related to those of the Guadalupe Mountains. To this group belong the species cited as P. semireticulatus, P. sinuatus?, P. costatus, P. mexicanus, P. plicatilis, and P. longispina; but Kayser's figures show that he included under some of these titles what I would regard as several species. Some of the larger forms, like P. semireticulatus and P. costatus, are more or less closely related to P. semireticulatus var. capitanensis and Productus sp. c, while the group of P. popei resembles in the Chinese fauna P. mexicanus and P. longispina. I do not know if P. plicatilis is correctly identified, but the type called by that name does not occur in the American fauna. The dorsal valve of P. waagenianus somewhat suggests the, Chinese shell.

The form which Kayser calls P. cora is clearly not the common Pennsylvanian shell which we are accustomed, and rightly I believe, to refer to D'Orbigny's species. It will subsequently appear that the Chinese shell is really of a different type, without any corresponding Guadalupian form.

Under the name of P. aculeatus Martin, Kayser figures what I would regard as several species, exhibiting characters that might warrant assigning some of them to the caperati, where De Koninck places P. aculeatus. These forms are comparable to the Guadalupian species which I have described as P. subhorridus var. rugatulus and P. walcottianus, perhaps even to P. latidorsatus, but not to the single Guadalupian representative of the caperati, P. pileolus. The American species, it will be remembered, I have placed with the horridi, doubtfully in the case of P. walcottianus, but the distinction between the caperati and the horridi is an artificial one. Another Chinese species which should perhaps be placed close to P. aculeatus is that described as P. kiangsiensis, but as Kayser's figures show a distinct or even rather high area it seems possible that this form may be a Strophalosia.

The shell described as P. pustulosus var. palliatus seems to be one of the fimbriati, a group whose absence from the Guadalupian has already been noted. P. nystianus var. lopingensis may possibly also belong to this group, though the figure suggests that the species may really be a Marginifera. The typical P. nystianus is one of the Proboscidei. Kayser also figures a form which he identifies as P. undatus, a type thus far unknown in the Guadalupian. The same is true of the form called P. carrangtonensis?.

The Lo Ping fauna has been revised by Fliegel and others, with numerous changes in the nomenclature and the specific divisions, but these affect the foregoing comparisons in but one instance. The form which, under the name of Productus cf. cora, Kayser figures so that it somewhat resembles P. pinniformis of the present work, is, according to Fliegel, the same as that which Kayser called P. undatus, for which Diener has proposed the name P. mongolicus. So interpreted it is of course very different from P. pinniformis, and indeed from any Guadalupian species yet found. On the whole, while there is a broad resemblance between the Productus fauna of the Guadalupe Mountains and that of Lo Ping in some particulars, the differences are perhaps still more marked.

As might be expected, this genus plays an important part in several of the faunas which Loczy described from different points in China. Nine species are cited from the vicinity of Kantschoufu, only one of which probably, the ubiquitous P. semireticulatus, has a closely allied form in the Guadalupian. In fact, it is probable that the Chinese fauna represents an older stage in the Carboniferous. Some of the species cited by Loczy belong to groups which are apparently absent from the Guadalupian, e. g., the undati and the fimbriati. A distinctly greater agreement, at least in the absence of forms alien to the Guadalupian, is shown by the Permian fauna from the vicinity of Batang. Three small species of the semireticulatus type were found there, one said to be allied to P. gratiosus, another to P. semireticulatus, and the third unidentified. These appear to be more or less like P. popei and its allies, or like P. mexicanus, while the only remaining form, which is cited as Productus cf. ovalis Waagen, somewhat recalls Productus sp. e.

No very close relationship is shown by the five species from the valley of the Lantsankiang. A figure of a fragmentary example cited as Productus aff. pustulosus is suggestive of P. signatus, though the latter is certainly not very closely related to Phillips's species. Another shell compared with P. seabriculus belongs to a type not found in the Guadalupian, but Loczy suggests that his specimen may really be an Aulosteges. The Productus semireticulatus is more or less like its Guadalupian congeners (perhaps Productus sp. c especially), but a little shell referred to P. tumidus appears not to be represented by any form in the American fauna. An entirely unidentified specimen,a however, is rather suggestive of P. walcottianus, although a more complete knowledge might lessen the resemblance.

---

aWissenschaftliche Ergebnisse der Reise des Grafen Béla Széchenyi in Ostasien, Wien, 1890, pl. 4, fig. 7.

---

Only two species are cited from the "Permo-Carboniferous" of Tschungtjen, in the province of Yunnan. One of these, referred to P. semireticulatus, a small, finely ribbed variety, is a little like P. semireticulatus var. capitanensis of the Guadalupe Mountains. The other is identified as P. aculeatus, and recalls two widely different types of the Guadalupian, P. mexicanus and P. subhorridus var. rugatulus. The lack of exact data in the description and figures of the Chinese form prevents a decision as to which group it represents, without reference to the original specimen.

But two species were distinguished from Talischau. The small finely ribbed Productus of the semireticulatus type similar to the foregoing is only in a general way like P. semireticulatus var. capitanensis or Productus sp. c. A little shell cited merely as Productus sp. is apparently related to P. latidorsatus and Productus sp. e.

The fauna obtained near Youngtschangfu appears to show little resemblance to the Guadalupian, in which there is nothing at all closely related to the three species of Productus cited from that locality.

In the Carboniferous fauna which Roemer described from Sumatra he discriminates five species of Productus—P. sumatrensis, P. pustulosus, P. cora, P. longispinus, and P. keyserlingianus. But one of these species is accompanied by figures, and so far as one may judge without this accessory they do not constitute a Productus fauna very similar to that of the Guadalupian. The same conclusion must be reached from a consideration of the representation of the genus in practically the same fauna as revised by Fliegel. This author cites P. lineatus (P. cora of Roemer), P. sumatrensis, P. semireticulatus, P. longispinus, P. ovalis (P. keyserlingianus of Roemer), and P. punctatus (P. pustulosus of Roemer).

Beyrich cited two species of Productus from the Carboniferous of Timor—P. semireticulatus and P. punctatus? There really appear to be three species represented by Beyrich's figures, none of which can properly be referred to either P. semireticulatus or P. punctatus, if the latter are reasonably restricted. One of the shells figured as P. semireticulatus appears to be rather closely related to P. texanus of this work, but the others are without Guadalupian representatives.

Martin seems not to have found the genus in the collections from Timor studied by him, but Rothpletz cites six species in his work on the Permian fauna of Timor and Rotti. Two of these, P. abichi and P. waageni, belong to the fimbriati, a group which, as I have several times had occasion to remark, is not found in the Guadalupian. The same is true of P. asperulus, which has no very close ally in that fauna. The semireticulati include the three remaining species of Productus recognized by Rothpletz—P. semireticulatus, P. gratiosus, and Productus n. sp. The resemblance of P. gratiosus to the group of P. popei and to P. mexicanus has already been commented on in connection with faunas from India, and the type here receiving that identification (P. gratiosus) is without doubt the same as that identified by Beyrich as P. semireticulatus, whose similarity to P. texanus has also already been remarked.

De Koninck recognized 12 species of Productus in his discussion of the Carboniferous faunas of New South Wales. None shows many analogies with the Guadalupian Producti, and indeed most of them occur in the lower series of the Carboniferous, and may well be regarded as beyond the bounds of the present discussion. A few species occur in both divisions, as, for instance, P. clarkei, which is cited both from Burragood, one of the lower horizons, and from Branxton, which, as it is not recorded in the list of localities where the lower series occurs, may be regarded as belonging to the upper.

Of the Producti which were obtained from the upper horizons P. clarkei, P. brachythærus, and P. undatus are perhaps the most important. P. clarkei was originally described as a Strophalosia, and has been considered elsewhere. De Koninck, according to his own account, saw no specimens of this species, and his assignment of it to Productus can not, therefore, be regarded as trustworthy. I repeat that I see but the most elementary resemblance between the Producti described by De Koninck and those of the Guadalupian fauna.

The genus Productus is represented in the "Permo-Carboniferous" of Queensland and New Guinea by something like 12 species, as discriminated by Etheridge, but a large number of these are not identified, and many are not figured in his report. Productus cora is not found in the Guadalupian, and in fact Etheridge's figures from Australia do not seem to me to belong to that species. The one most nearly resembling P. cora is a dorsal valve from the Mount Britton gold field. Productus brachythærus, P. subquadratus, and P. undatus are non-Guadalupian, though the figure of the latter, said to be a ventral valve, suggests the dorsal valve of P. waagenianus. P. semireticulatus is not figured, but presumably resembles P. semireticulatus var. capitanensis and Productus sp. c. P. longispinus? is more or less semblable to the Guadalupian species of the group of P. popei. The remaining Producti are unidentified, and with one or two exceptions appear to belong to non-Guadalupian types. Productus sp. d and Productus sp. f, however, are probably related to the Guadalupian shells of the group of Productus popei, the former more to P. popei itself and the latter to P. texanus.

In the Russian section the genus Productus shows a remarkable differentiation, reaching its acme in the Gschelian, where Tschernyschew recognizes the extraordinary number of 43 species. In the Artinskian the number is much less, while in the Permian I have found recorded but five. It would protract this discussion needlessly to compare the Guadalupian Producti with those of the Productus giganteus and Spirifer mosquensis faunas, but I may remark in passing how much the Productus fauna of the Moskovian resembles that of our Mississippi Valley Pennsylvanian.

Tschernyschew divides his 43 species of Gschelian Producti into 20 groups, the first of which, the group of P. boliviensis, comprises five species of the semireticulatus type. This group of forms is represented in the Guadalupian only in a general way. Nearest perhaps are P. guadalupensis, which resembles P. multistriatus Meek var. of Tschernyschew's report, and Productus sp. c., which resembles P. boliviensis and P. gruenewaldti. As to the identity of the Russian species with our American form, the resemblance is certainly very close, although Meek does not mention or figure the visceral area as being marked by concentric corrugations, the absence of which would preclude the assignment of this form to the semireticulati. Rather faint concentric folds are shown in Tschernyschew's figure. It should be noted that P. multistriatus is a distinctly western form in the American Carboniferous faunas, no corresponding type, so far as I recall, being found in the Pennsylvanian of the Mississippi Valley. I have been tentatively correlating the horizon of P. multistriatus with beds below the Guadalupian.

The next group, that of P. semireticulatus itself, has three species. Like the foregoing, this group finds representation in a general way in the Guadalupian fauna, in this case especially by P. semireticulatus var. capitanensis. One of the members of this group Tschernyschew identifies with our little-known American species P. inflatus. Among the most typical semireticulati there is so strong a common resemblance and so many intermediate forms that it is difficult to determine and maintain specific limits. I have not seen specimens which could without hesitation be placed with P. inflatus, and feel doubtful as to the reference of the Russian specimens.

The third group comprises a single species, Productus mölleri, one of the semireticulati of the usual type.

All these species are more or less closely related to Productus semireticulatus var. capitanensis, Productus sp. c, and P. guadalupensis.

The fourth group is represented by P. lobatus and P. mexicanus White (Shumard?). These two forms are small species of the semireticulatus group, a type which is well represented in the Guadalupian. The Guadalupian forms, however, all have a deep sinus, while the Russian species are represented without this feature. P. mexicanus was described from the "white limestone" of the Guadalupe Mountains, and unfortunately I have been unable to identify it among my material. This is the more to be regretted since Shumard's description was not accompanied by figures. I am doubtful as to the correctness of White's identification. His species is considerably smaller, and its occurrence appears to be at a lower geologic horizon, since I am tentatively assigning it to a position below the Guadalupian. Tschernyschew's Russian specimens certainly resemble White's identification. It will be remembered that this species has also been identified in China. P. mexicanus in North America is distinctly western. So far as I can recall, it has no closely allied representatives in the Pennsylvanian faunas of the Mlississippi and Ohio valleys. Productus sp. a. resembles the Russian shell figured as P. mexicanus White (Shumard?).

The next group, comprising only P. tartaricus, represents the type which has so many Guadalupian representatives, among which P. popei and P. texanus are conspicuous.

The sixth group, that of P. stuckenbergi, contains but P. stuckenbergi itself, another small species similar to P. popei and P. texanus, but having a strongly corrugated visceral area.

With these the division of the semireticulati comes to an end, although it apparently is resumed at a later point. In a general way it may be said that the Russian series of forms resembles the Guadalupian rather closely, but shows a greater differentiation of large species closely related to the typical P. semireticulatus. The absence of species of the type of P. occidentalis from the Russian faunas may also be noted.

The group of P. spinulosus (No. 7) comprises seven species. These seem to have Guadalupian representatives in P. latidorsatus and P. subhorridus var. rugatulus. The resemblance in some cases is rather strong, as between P. subhorridus var. rugatulus and the Russian shell identified as P. wallacianus, and between P. latidorsatus and P. tastubensis, while some of the Russian forms have no very analogous types in the Guadalupian, as, for example, P. pustulatus Keyserling.

The next group, that of P. humboldti, has two Russian representatives and none at all in the Guadalupian. The same is true of the ninth group, called that of P. nebraskensis. One of the two forms is doubtfully referred to our common American species P. nebraskensis. The Russian specimen figured seems to be so poor that the identification may well have been held in doubt.

The next group includes representatives of the striati to the number of three. The shell figured as P. cora certainly seems to be the same form which D'Orbigny described from South America and which is abundant in the Pennsylvanian faunas of the Mississippi Valley. This type, however, does not occur characteristically in the Guadalupian. The little shell identified by Tschernyschew as P. aagardi Toula is suggestive of P. waagenianus, but the dorsal valve appears to lack the peculiar configuration of the American species.

The eleventh group is composed of three species, two of which appear to be closely related to Productus cora. The third is less so, but none has a closely comparable form in the Guadalupian.

The next two groups, each with a single species, represent types more or less closely related to P. cancrini, and find representation in the Guadalupian in P. meekanus and P. signatus.

Productus pseudomedusa, the singular species on which Tschernyschew's fourteenth group is based, has no corresponding form in the Guadalupian fauna.

In P. artiensis and P. mammatus, which represent the fifteenth group, we again find species related to Guadalupian forms, the former to P. popei, the latter more or less to Productus guadalupensis, or possibly to Productus sp. d.

The group of P. punctatus (No. 16) comprises two species, neither of which has any related type in the Guadalupian. It is perhaps worthy of note that Tschernyschew's identification of P. punctatus is quite unlike the American form commonly referred to this species, which, however, is very similar to P. fasciatus, the other member of the group of P. punctatus in the Russian work.

The seventeenth and eighteenth groups, the former with one and the latter with two species, have no representatives in the Guadalupian. The corresponding American species occur at horizons which I regard as older than the Guadalupian and associated with a very different fauna. These are western forms, without closely allied representation in the Pennsylvanian of the Mississippi Valley. The shells which Tschernyschew figures under the name of P. longus are so different from Meek's figure of that species that the identification must be regarded as at least doubtful.

P. timanicus, which alone constitutes the nineteenth group, finds a representative in the Guadalupian fauna in the related P. latidorsatus. The resemblance is not very striking, consisting more in sculpture than in configuration, and mature specimens are rather strongly different.

The twentieth group comprises but two species, P. anomalus Keyserling, whose Guadalupian representative is P. pinniformis, and P. ischmensis, which has no corresponding species in the American fauna.

To take the Producti as a whole the Guadalupian fauna, in paradoxical language, may be said to resemble the Gschelian more closely than the Gschelian does the Guadalupian. In other words, while most of the Guadalupian species find Gschelian types more or less closely related, the latter fauna contains much that is unrepresented in the Guadalupian. Some of these forms may be regarded as survivors from earlier faunas, as, for instance, species closely related to P. punctatus, P. cora, and P. nebraskensis. In many cases these non-Guadalupian species have similar or identical American representatives, but as a rule the latter occur in our western areas and in association with faunas markedly different from the Guadalupian, and found, I believe, at distinctly lower horizons.

Tschernyschew's account may fairly be taken as giving a representative exposition of the Gschelian Productus fauna, and while other reports in which this fauna has been discussed have come before me, they can well be neglected in the present discussion, whose scope is more general than particular.

In attempting to ascertain the character of the Artinsk fauna so as to compare with it that from the Guadalupe Mountains, instead of a single volume in which the species are all figured, as in Tschernyschew's work on the Gschelian, I have had to consult several in which the fauna was for the most part listed. The largest Productus fauna in the Artinsk seems to be that discussed by Stuckenberg, who lists 21 species. In spite of the fact that but four of these are figured, it is possible to point out that a considerable number have allied species in the Guadalupian. Many of these either occur in the Gschelian or are represented there by cognate forms. Seven of the Artinskian species belong to the semireticulati. P. semireticutalus, P. boliviensis, Productus cf. spiralis, and P. möelleri are large types more or less closely related to P. semireticulatus var. capitanensis and Productus sp. c in the Guadalupian, while P. longispinus, P. stuckenbergianus, and Productus cf. artiensis are small species more or less comparable to P. texanus and P. popei. Productus cora and the related P. tenuistriatus are not represented in the Guadalupian by forms closely allied to them. P. koninckianus and P. cancrini find allied species in P. meekanus and P. signatus. The following have no related Guadalupian species and in some cases none belonging to the same section: P. scabriculus, P. punctatus, P. silvæanus, P. fimbriatus, and P. granulosus. P. aculeatus may have a somewhat distantly connected form in P. walcottianus. P. tuberculatus rather suggests P. latidorsatus, with which P. timanicus may also be in some respects compared. P. krasnopolskyanus presents an appearance not unlike P. subhorridus var. rugatulus.

How closely related are the Productus faunas of the Gschelian and the Artinskian appears from the fact that of the 20 species which Stuckenberg cites from the Gschelian 13 run up into the Artinskian, while of the 21 from the Artinskian the same number of course come up through the Gschelian. So far as this report is concerned, therefore, practically the same remarks which were made regarding the Gschelian may be repeated of the Artinskian Producti in their relation to those of the Guadalupe Mountains. Most of the Guadalupian species have forms in the Artinskian to which a correspondence can be traced, intimate in some cases but remote in others, while in the Artinskian are a number of types which have allied forms in the Hueco formation and correlated beds but none in the overlying Guadalupian.

Tschernyschew also gives a list of Artinskian Brachiopoda, many of them figured, in which he cites 12 species representing about the same type of fauna as Stuckenberg's list, many of the species being identical.

Krotow also lists a large number of species from the Artinsk sandstone, only a few of them, unfortunately, being figured. The list comprises 23 species, four of them unidentified, and includes eight which Tschernyschew identifies in the Gschelian. Among the types found by Krotow are large species of the semireticulatus group, as well as small ones several of which are closely allied to Guadalupian species such as P. popei, P. texanus, P. indentatus, etc. The Russian forms especially in mind are P. longispinus and P. stuckenbergianus. There are also P. koninckianus and P. cancrini, which appear to be related to P. meekanus and P. signatus. On the other hand Krotow cites P. cora, together with a number of the fimbriati, such as P. humboldti, P. fimbriatus, P. punctatus, a group whose absence from the Guadalupian has already called for comment.

From the Permian Stuckenberg obtained but few invertebrates and no brachiopods, but Netschajew cites four Producti—P. cancrini Vern., P. hemisphærium Kut., P. hemisphæroidalis Stuck., and P. planohemisphærium Stuck. P. cancrini finds related species in the Guadalupian in P. meekanus and P. signatus, but there are no Guadalupian forms closely resembling the three others. About the same species (P. cancrini and P. hemisphærium) are recorded from the Permian by Golowkinsky, while from the province of Kostroma Tschernyschew cites only P. cancrini, but his figures represent a form which I would think more nearly related to P. cora. It appears to lack the regular distribution and elongate bases of the spines, which are a striking character of De Verneuil's figures. Sibirzew cites P. cancrini and the related Productus aff. koninckianus from the lower Permian series and P. cancrini from the upper, while in the original work on the Permian De Verneuil describes P. cancrini and P. leplayi, the latter being one of the semireticulati.

From these data it would appear that there is a great difference between the Producti of the Russian Permian and of the Artinsk, manifested, however, more in the defection of old types, with a general falling off in the representation, rather than in the introduction of new ones. Exception may perhaps be found in the three species recognized by Netschajew, but while they are poorly figured and apparently rather characterless types, I believe that they are to be considered rather survivors than newly introduced. It is hardly necessary to comment on the Guadalupian Producti in this connection. They contain much that the Russian Permian does not, but the donminating types of the greatly diminished Productus fauna of the latter are in the main to be found in the Guadalupian.

The fauna described by Enderle from Balia Maaden, in Asia Minor, need not long detain this discussion, though it contains, according to this author, Producti to the number of 21. Four of these are placed with the lineati—P. lineatus, Productus cf. cora, Productus cf. margaritaceus, and P. mysius. To the two latter especially the Guadalupian fauna contains no corresponding forms, although Productus cf. margaritaceus seems to me to be referred to the lineati with doubtful propriety and to be more probably a poorly characterized type of the semireticulati. To the two former, though P. lineatus is not figured, the Guadalupian species P. waagenianus is somewhat distantly related.

Nine species represent the semireticulati, according to this author, over against which there is about an equal number of Guadalupian varieties. In a general way the semireticulati of the two faunas present many correspondences. It could hardly be otherwise; but the large types are prevalent in the fauna from Balia Maaden and the small ones in that from the Guadalupe Mountains. The differences are not very marked, the latter fauna, for instance, lacking a corresponding form to P. semireticulatus var. bathykolpus, and the former being without anything closely related to P. guadalupensis and P. occidentalis.

To the Proboscidei Enderle refers a shell identified as P. nystianus De Kon., though I can not but doubt profoundly either the direct reference to the species or the implied one to the genus Proboscidella, provided the figure be correct. The form in question appears in fact to be very similar to that called "Productus cf. margaritaceus," both types being more probably small examples of the semireticulatus section, related, though not necessarily very closely, to P. texanus and its allies. "Productus aff. undati" (unfigured) finds a related form in the Guadalupian, if at all, in P. meekanus. P. punctatus, representing the fimbriati; P. scabriculus, representing the spinosi; and Productus cf. tumidus, incorrectly, I believe, referred to the horridi, have no related Guadalupian species. The same is true of P. aculeatus?, referred to the caperati and not figured, unless Productus pileolus or P. subhorridus var. rugatulus prove to resemble it. P. subhorridus var. rugatulus is also comparable in some ways to the little shell which Enderle identifies as P. curvirostris, but P. pileolus is far more closely related. The third species referred to the caperati, P. troianus, is quite unlike any Guadalupian types of Productus. This unusual form in fact rather suggests some species of Aulosteges. On the whole the Producti of the Balia Maaden fauna show no very close relationship with the Guadalupian species.

Much more nearly related in some respects is the fauna from Armenia described by Abich. The most striking feature of the Producti of this fauna is the great development of species having a general resemblance to P. latidorsatus, P. subhorridus var. rugatulus, and P. walcottianus. Abich cites members of this group under different names, but one would be disposed to refer to it the forms called P. intermedius, P. intermedius var. planiconvexus, P. intermedius var. helicus, P. spinosicostatus, P. spinosicostatus var. cariniferus, P. spinosicostatus var. expansus, P. spinosi costatus var. incurvus, besides P. martini, P. aculeatus, and P. spinulosus. These appear to be related, some of them very closely and others remotely, to the three Guadalupian species mentioned above, but unfortunately it has subsequently been shown that all except P. intermedius, P. intermedius var. planiconvexus, and P. martini belong to Waagen's genus Marginifera, the structures characterizing which have not been observed in the Guadalupian species and appear in some cases to be absent. If, therefore, we except these forms, in which the resemblance, it would seem, is only apparent the relationship between the Armenian and the American faunas in point of the genus Productus is not very close.

Abich regards P. intermedius as a member of the semireticulati and Arthaber who subsequently reworked the fauna, takes the same position, even referring to the same species P. intermedius var. planiconvexus Abich and P. martini Abich non Sow.; but I can not see any justification for this, for while the visceral area is marked by concentric corrugations, both authors represent the form in question as entirely without ribs. In my view they should be placed with the horridi or caperati, although these groups are described as typically without concentric corrugations. They appear to resemble, in a general way, the Guadalupian species which I have called "P. latidorsatus." With the foregoing exception the great group of the semireticulati would appear to be unrepresented in the Armenian fauna, a really surprising circumstance, since these shells are seldom absent where the genus Productus occurs at all.

The remaining Armenian species of Productus belong to the lineati and the fimbriati. To the latter, a group, it will be remembered, which does not occur in the Guadalupe Mountains, are referred P. scabriculus, to which as identified by Abich. Waagen subsequently gave the name P. abichi, and P. humboldti, for which, in like manner, the name P. waageni was later substituted by Rothpletz. The lineati comprise three species which Arthaber refers to P. hemisphærium Kut. (P. hemisphærium var. armeniacus in the description of plates), but which Abich distinguishes as P. striatus Fischer, P. striatus var. sphæricus, and P. undatus. Some of Abich's figures, especially that of P. striatus, are very suggestive of our Productus cora, and one would be inclined to think that Abich may have two species, instead of one. Arthaber places these forms with the irregulares, but they seem to belong more properly in the lineati. If they have any Guadalupian representative it is P. waagenianus or P. waagenianus var.

Arthaber recently redescribed Abich's fauna, introducing many changes, some of which have already been noted. He also introduced one new specific name, that of P. mytiloides, of which P. pinniformis would probably be the Guadalupian representative. On the whole, if the group of shells which Arthaber withdraws to the genus Marginifera is eliminated, the Armenian fauna from Djoulfa does not, in its Producti, show any marked resemblance to the Guadalupian.

Gemmellaro's report on the Producti of the Sicilian fauna from Palermo being, unfortunately, inaccessible, I must next examine those by Schellwien on collections from the Carnic Alps. In his paper on the fauna of the Fusulina limestone this author cites ten species. Three of these belong to the lineati—P. lineatus, P. cora, and P. cancriniformis. The first is perhaps rather remotely related to P. waagenius, but it is doubtful if P. meekanus should be considered as the representative of P. cancriniformis. The semireticulati, comprising P. semireticulatus, P. semireticulatus var. bathykolpus, P. gratiosus var. occidentalis, and P. longispinus, have allied forms in the Guadalupian, P. semireticulatus var. bathykolpus in P. semireticulatus var. capitanensis and P. gratiosus var. occidentalis, and P. longispinus in Productus sp. c, P. popei, P. texanus, and cognate species. P. punctatus, however, has no corresponding type in the Guadalupian, but P. aculeatus var. may possibly be compared with P. walcottianus and P. subhorridus var. rugatulus. These forms also resemble in a general way P. curvirostris, to which, however, a little shell that may possibly belong to the genus Marginifera has a still closer superficial resemblance. On the whole Schellwien's fauna seems in its Productus content to be rather similar to the Guadalupian, though the latter contains some types which do not occur in the Alpine fauna (e. g., P. occidentalis and P. latidorsatus).

In the fauna of the Trogkofelschichten the same author cites 14 species, in many cases the same as those of the preceding one. The group of P. cora contains only P. cora itself, distantly related to the Guadalupian species P. waagenianus. The group of P. cancrini consists of P. cancriniformis and P. cancriniformis var. sinuatus, the former comparable to P. meekanus and P. signatus. The group of P. semireticulatus comprises P. semireticulatus, P. semireticulatus var. bathykolpus, and Productus cf. spiralis. They are more or less closely related to Productus semireticulatus var. capitanensis and Productus sp. e. The group of P. griffithianus is represented by P. gratiosus, which save for its strongly wrinkled visceral area is very suggestive of the group of P. popei. P. aculeatus, representing the group of that name, is related to P. walcottianus. P. spinulosus, P. tuberculatus, and Productus sp., belonging to the group of P. spinulosus, are similar, though not very closely similar, to P. latidorsatus, P. subhorridus var. rugatulus, and their allies. P. curvirostris, somewhat related to the same species, is much more similar to a little shell described by Shumard as Productus pileolus, which is possibly to be referred to Marginifera. P. elegans and P. incisus have no Guadalupian species closely allied to them. The Guadalupian Producti, however, in their general facies are rather comparable to those of the Trogkofelschichten.

Gortani also described a fauna from the Carnic Alps, but one which I take to be older than that of the Trogkofelschichten. Fifteen species are distinguished by this author. The three varieties of the cora group have, so far as known, no immediate relatives in the Guadalupian. One species is referred, somewhat doubtfully, to P. giganteus, a type which would hardly be looked for in this faunal association. The figures suggest a cast of the dorsal valve of Derbya or some other strophomenoid. The larger semireticulati are identified as P. semireticulatus, P. semireticulatus var. transversalis, and P. semireticulatus var. bathykolpus. These correspond in a general way to P. semireticulatus var. capitanensis, Productus sp. c, and P. guadalupensis of the Guadalupian. The smaller types, such as P. gratiosus, P. longispina, and P. longispinus var. lobatus, though not figured, are probably comparable to P. popei and its allies. It is true, however, that Gortani cites P. longispina as a Marginifera, though Waagen has stated that typical P. longispina does not belong to that genus. The remaining species, P. punctatus, Productus cf. fasciatus, P. elegans, P. humboldti, and P. abichi, appear to have no Guadalupian allies, and indeed the whole Productus fauna appears to have a different facies.

In the Dyas of Germany Geinitz recognizes five species of Productus—P. cancrini, represented in the Guadalupian by P. meekanus and P. sigatus; P. hemisphærium, perhaps not properly belonging in the fauna and without closely related Guadalupian species; P. latirostratus, possibly an Aulosteges and without a Guadalupian representative; P. horridus, appearing to be represented in the Guadalupian by the not very closely related species P. latidorsatus; and P. geinitzianus, possibly to be correlated with the form just mentioned, though in some respects comparable also to P. walcottianus and to P. occidentalis. Perhaps I should also include a shell placed by Geinitz in the genus Strophalosia and identified with De Verneuil's species Productus leplayi, which I believe is by its author correctly regarded as a Productus. By way of illustration Geinitz produces a figure copied from De Koninck but evidently originally derived from De Verneuil's illustration of Productus leplayi.

While comparable in some respects the Producti of the Dyas fauna are specifically much less varied than those of the Guadalupian. Except for P. leplayi, above mentioned, we miss the great group of the semireticulati, and especially the small, arched, deeply sinused forms of the Guadalupian fauna. We miss also the cora group and the irregulares of Waagen. The absence of other types, such as the fimbriati, is shared by the Guadalupian fauna.

In the Permian of England, closely related to that of Germany, King recognizes but one species of Productus (P. horridus), P. umbonillatus, as King suggests, being probably an Aulosteges.

From Spitzbergen De Koninck cites four species of Productus—P. horridus, P. leplayi, P. cancrini, and P. robertianus. The first of these, unless an imperfect specimen or figured very badly, is not a characteristic example of the species to which it is referred. It is not very unlike P. latidorsatus or even P. subhorridus var. rugatulus. P. cancrini and P. leplayi correspond in a general way to the Guadalupian forms P. meekanus or P. signatus, and P. semireticulatus var. capitanensis or Productus sp. c. The little shell called P. robertianus is certainly not like any ordinary type of Productus. In fact, from the figures one might be excused for mistaking it for a Spirifer I suspect, however, that it will be found to be a Chonetes, belonging to the group of the grandicostati.

In his paper on fossils from the south point of Spitzbergen Toula cites six species of Productus. P. payeri and P. weyprechti probably, and P. humboldti certainly, have, so far as known, no Guadalupian representatives. The shell referred to P. koninckianus is somewhat similar to P. meekanus and P. signatus. The two undetermined species are not figured and can not enter into a comparison.

In describing later a fauna from the Hornsund, on the west coast of Spitzbergen, this author cites seven species of Productus, to which should perhaps be added a fragmentary specimen identified as P. leplayi De Vern and referred to the genus Strophalosia. P. weyprechti, which in the preceding paper had been figured from an internal mold, is here shown to be one of the semireticulati, together with P. spitzbergianus, which Toula, I can not but think with little reason, compares with P. horridus. P. weyprechti seems to be most nearly related to P. guadalupensis of our American fauna. Another of the semireticulati, P. wilczeki, strikingly resembles P. popei and its allies, some of which are comparable to the shell that Toula identifies as P. longispinus, perhaps also to the unfigured P. longispinus var. acutirostratus. The form referred to Productus cf. prattenianus (= P. cora) is said to be without spines, and therefore it is perhaps more nearly allied to P. lineatus. There is no Guadalupian species very similar to it. The form identified as P. undatus is not figured. If it is the same as typical P. undatus, no species in the Guadalupian can be compared to it, except, very loosely, P. meekanus.

From Axel Island Toula cites three varieties of P. horridus, P. cancrini, Productus cf. humboldti, P. weyprechti, Productus sp., P. semireticulatus, P. aagardi, and P. impressus. Of the modifications of P. horridus one a has much the configuration of P. guadalupensis and is represented as being marked, especially toward the front, by obscure ribs. The other forms appear not to be striated and can be compared, though but partially, with P. latidorsatus and P. subhorridus var. rugatulus. P. impressus, Productus cf. humboldti, and P. weyprechti appear to be without Guadalupian representatives. P. cancrini (not figured) is perhaps analogous to P. meekanus and P. signatus, P. semireticulatus to P. semireticulatus var. capitanensis and Productus sp. c, though not very similar, and P. aagardi to P. waagenianus.

---

aToula, F., Neues Jahrb., 1870, pl. 5, fig. 2.

---

From the cape between the two arms of the North Fjord Toula cites P. horridus var. spintzbergianus, P. cancrini, P. humboldti, Productus cf. scabriculus, and P. longispinus var. setosus. P. horridus var. spitzbergianus seems to be the same species which this author described from the Hornsund as P. spitzbergianus and from Axel Island as P. horridus. It is possibly related to P. guadalupensis. The form referred to P. cancrini is perhaps related to P. meekanus and P. signatus, though Toula's specimen appears to have been in an imperfect state of preservation. Possibly the same cause would account for differences between Toula's figures and those of typical P. scabriculus. This form, together with P. humboldti and P. longispinus var. setosus, appears to have no closely related species in the Guadalupian. The last named is possibly the species which De Koninck described from Spitzbergen as P. robertianus, and if so my surmise that P. robertinanus is one of the grandicostate Chonetes must be in error. It more nearly resembles P. popei and related forms than any other Guadalupian species.

Lundgren cites only Productus ? sp. as occurring in his Permian fauna from Spitzbergen, and it may be said in general, though with here and there an exception, that the Producti of the Spitzbergen faunas are not very similar to the Guadalupian ones.

This is still more true of the fauna from the Barents-Inseln, Nova Zembla, from which Toula cites P. cora, P. semireticulatus, P. costatus, P. punctatus, P. humboldti, P. aculeatus, and P. obscurus. P. obscurus and P. aculeatus may in a general way correspond to P. subhorridus var. rugatulus and P. latidorsatus, and P. semireticulatus and P. costatus to P. semireticulatus var. capitanensis and Productus sp. c. I can not repress a suspicion that Toula's figure of Strophomena depressa represents a dorsal valve of a Productus of the cora type, rather than a Leptæna.

From the Productus limestone of the Wadi-Draa in the West Sahara Stache has brought to notice a remarkable Productus fauna, consisting of numerous mutations of what appear from figures to be a single primal form. They all show more or less close relationship to P. hemisphærium and P. margaritaceus, the great groups of the semireticulati, fimbriati, etc., being entirely unrepresented. One species, it is true, P. devestitus, Stache refers to the group of P. sublævis, but it is certainly not a normal representative of the semireticulati. In the Guadalupian the only species which may be compared with these is P. waagenianus, and the relationship is probably not close. Stache distinguishes 13 species in this series, only two of which need occasion further remark. One of the new names introduced by Stache is P. semistriatus, the combination used by Meek for an American shell in 1860. The form called Productus? tripartitus, both from the figures and the structures which it is said to possess, I suspect to belong to another genus—to be, in fact, a dorsal valve of a strophomenoid.

Stache also cites two species of Productus (Productus aff. margaritaceus and Productus sp.) from the "Sandsteinschichten der Mittelregion" of the West Sahara and four from Igidi; but as these forms have little to do with the Guadalupian fauna I shall not delay over them, pausing only to comment on numerous mistakes in plate references which occur in this work and cause labor and uncertainty to anyone wishing to consult them. It seems as if examples of all possible errors could be furnished in text, plates, and descriptions of plates, including incorrect citation, duplication, and omission of numbers.

From Bolivia Toula cites only Productus cf. cora and P. semireticulatus. Salter cites only P. semireticulatus and P. longispinus (P. capacii of D'Orbigny), the latter very similar to P. texanus of the Guadalupian. In D'Orbigny's volume, however, we find nine species described. One can easily trust too far to D'Orbigny's figures, but it is safe to say that no species of the type of P. humboldti are as yet known in the Guadalupian. P. villiersi appears to correspond to P. meekanus. P. cora is without a closely corresponding form. P. andii has been shown by De Koninck to be an orthoid (= Orthis buchii). P. capacii is closely similar to P. texanus, and P. inca, P. peruvianus (fide De Koninck), P. boliviensis, and P. gaudryi (figured but not described) are large species of the semireticulatus type more or less allied to P. semireticulatus var. capitanensis and Productus sp. c. Upon the whole the known Peruvian faunas are not very similar to those of the Guadalupian in point of the genus Productus, and I am tentatively holding that they represent an earlier epoch.

In his Peruvian material Gabb distinguishes only three species of Productus, one a small form neither described nor figured, the two others large species belonging, it would appear, to the semireticulati.

The Brazilian fauna which Derby described includes seven species of Productus. P. semireticulatus, P. chandlessii, P. batesianus, and P. rhomianus (if the latter does not after all belong to the genus Marginifera) can all be referred to the semireticulati, without, however, being very similar in any instance to the Guadalupian species P. semireticulatus var. capitanensis. They appear to be nearer to the imperfectly known Productus sp. c. P. chandlessii and P. batesianus, however, have cognate species in P. guadalupensis, and P. rhomianus possibly in P. texanus and its allies. P. cora has no very close ally in the North American fauna. P. clarkianus bears comparison with P. meekanus, and P. wallacianus to a certain extent with P. latidorsatus and P. subhorridus var. rugatulus. Though I am tentatively regarding the Brazilian fauna as older than the Guadalupian, the Producti show considerable resemblance.

In his valuable bibliography Weller recognizes upward of 50 species of Productus from the Upper Carboniferous rocks of North America. Fifteen of these are distinctly western forms, the remainder being for the most part restricted to the Mississippi Valley and Appalachian regions. To treat the Pennsylvanian Producti as I have attempted to do in the case of foreign faunas would be impossible, since many of them have not been figured; and the need of detailed comparisons is less imperative, since all who are familiar with the Pennsylvanian faunas will at once recognize that they are very unlike those of the Guadalupe Mountains. Suffice it to recall in this connection that the Producti of the Pennsylvanian commonly belong to five types which occur again and again, whether collections be made at different localities or different horizons. We have Marginiferas such as M. wabashensis and M. muricata, and species of Productus more or less closely related to P. sentireticulatus and P. costatus, to P. cora, to P. nebraskensis, and to P. punctatus. Of these the greatest variation is shown by the semireticulati. None of the Guadalupian species of Productus is, so far as I am aware, identical with a Pennsylvanian form.a Some of the semireticulati have analogous forms in the two faunas, but the punctatus and nebraskensis types are entirely alien to the Guadalupian, and P. cora nearly as much so, being represented only by the distantly connected P. pinniformis and P. waagenianus.

---

aShumard cites P. norwoodi from the Capitan formation, but I am at a loss to know with what he had to do. If not an Aulosteges the form in question would appear to be one of the fimbriati, a type which, at least in our collections, seems to he unrepresented in the Guadalupian.

---

Species of Marginifera more or less completely identical with M. wabashensis and M. muricata are a rather constant factor in most Pennsylvanian faunas. The genus Marginifera is, so far as known, entirely lacking from the Guadalupian.

Productus pileolus, the species which seems from its structure most likely to belong to that genus, is of an altogether distinct type from the Pennsylvanian Marginiferas. On the other hand, the forms which superficially most resemble the Pennsylvanian Marginiferas, such as Productus subhorridus var. rugatulus and certain shells belonging to the group of Productus popei, seem to be without the characteristic Marginifera structure.

Among the rarer Pennsylvanian Producti one at least seems to have a related type in the Guadalupian, P. pertenuis being unquestionably related to P. meekanus, but it can be stated without qualification that the Guadalupian Producti are markedly different from the Pennsylvanian forms. Nor will it be necessary to devote much time to the scattered species which have been described from the West. In no instance, I believe, are they identical with Guadalupian species, and the associated faunas in every case, so far as I am aware, are different and probably older.

In thus hastily surveying the representative of this most characteristic Carboniferous genus in the different faunas of the world it has become apparent that the Guadalupian in this particular is an individual entity. Its affinities with certain faunas are marked, but almost invariably it presents equally marked differences, so that it can not be said to be identical with any of them.

PRODUCTUS WAAGENIANUS n. sp.

Pl. XII, figs. 6 to 7a.

Shell small. Ventral valve much inrolled, gradually spreading transversely. Beak inflated, pointed, strongly incurved, slightly projecting. Ears small, depressed, quadrate, flattened. Sinus absent. Toward the margin the shell develops a few low folds.

Dorsal valve nearly planate, with a narrow geniculated portion around the front and sides. Beak small, indistinct. Ears undefined.

Surface marked by radiating liræ, concentric liræ, concentric wrinkles, and spines. Liræ very fine, about 14 in 5 mm.; low, rounded, separated by intervals as wide or wider than themselves. Concentric liræ rather coarse but indistinct. Wrinkles faint, distant, covering more than half the surface; strong on the ears. Visceral portion with a few rather large nodes, which may have been bases of spines, some of which appear to have been located on the ears, especially near the hinge line.

The foregoing description, so far as it relates to the surface, is based on the ventral valve. The surface of the dorsal valve is marked by concentric wrinkles (and presumably concentric liræ) and radiating liræ. The wrinkles on this valve are very strong, regular, subimbricating, and rather distant, covering the surface as far as the geniculation. The shape and ornamentation are such as to simulate certain varieties of Leptæna rhomboidalis. The wrinkles are so much stronger than those of the ventral valve as to suggest that the two shells do not belong together; but the other characters are similar, they are associated in the same beds, and nothing has thus far come to hand which in either case could be taken for the supplementary valve. I have at present little doubt about their relationship. Were it not for the peculiar character of the dorsal valve it might perhaps have been possible to refer this form to the common P. cora; but with the present association it is out of the question.

P. waagenianus is related to the Arctic species P. aagardi Toula not only in a general way but in the plicated condition of the dorsal valve. It is distinguished, however, by its much finer liration.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

PRODUCTUS WAAGENIANUS var.

The typical specimens of P. waagenianus were derived from the white limestone of the Guadalupe section. From the sandstones of the Delaware Mountain formation a few ventral valves of a similar type have been obtained, though I doubt if they can properly be referred to the same species. Nevertheless the fossils so far observed are really too poorly preserved to warrant the introduction of a new name. The convexity of these shells is variable, though generally broad and low. The ears are large, flattened, and undefined. The surface is marked by rather strong thin liræ, with rounded intermediate grooves. There appear to be no wrinkles, except faint ones on the ears, and a few small spines can be observed in the same region. The dorsal valve is not known.

The main points of difference from P. waagenianus consist in the coarseness of the liræ, in which the Delawarian specimens show some differences among them selves. The finest of them, however, has but eight or nine in the space of 5 mm., a distinctly coarser liration than in P. waagenianus proper. This form resembles P. cora, but even if the dorsal valve is not constructed as in the typical variety, the absence of large spines over the surface should serve to distinguish them.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (stations 2903 and 2919).

PRODUCTUS SEMIRETICULATUS var. CAPITANENSIS n. var.

Pl. XII, figs. 1 to 3b; Pl. XX, figs. 8 and 8a.

1858. Productus semireticulatus. Shumard (non Martin), Trans. Acad. Sci. St. Louis, vol. 1, p. 292 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

1859. Productus semireticulatus var. antiquatus. Shumard (non Martin), idem, p. 389.

White [Permian] limestone: Guadalupe Mountains.

Shell large. Ventral valve strongly arched. Beak moderately full. Ears probably large and projecting. Sinus profound, angular, and extending nearly to the beak.

Dorsal valve with strongly flattened visceral region; anterior and lateral portions uniting in an abrupt geniculation and at an angle somewhat less than 90°. Sinus faint over the visceral region, stronger below the geniculation. Ears large, defined by a groove, somewhat arched. Both valves exhibit a tendency to develop along the margin into large, loose folds and irregular growths.

The surface is marked by moderately fine ribs and regular concentric wrinkles of about the same size. The wrinkles are restricted to the visceral area, and produce regular nodes where they cross the ribs. Some of the nodes are more prominent than the rest, and formed the bases of small spines. The ribs increase by bifurcation, and those nearest the sinus exhibit a marked tendency to run down from the sides into it. They come about four or five in 5 mm. The distribution of the spines has not been ascertained.

While the dorsal valve, and doubtless the ventral one also, had much-produced ears, it is noticeable that the growth lines and wrinkles contract at the hinge. The sinus is remarkably deep and angular and reliance is placed on this and other peculiarities in the configuration and ornamentation to distinguish this form from the varieties of P. semireticulatus found in the Mississippi Valley, and also from the typical English species. This is undoubtedly the variety which Shumard mentioned in his original list of fossils from this locality, referring to it as Productus semireticulatus. His remarks are as follows:a

---

aTrans. Acad. Sci. St. Louis, Vol. 1, 1856-1860, p. 292.

---

This widely distributed species is contained in Captain Pope's collection from the white limestone of the Guadalupe Mountains. The specimens resemble most the variety E. antiquatus, but the sinus of the receiving valve is more profound and narrower than in the example figured by De Koninck, which are generally marked with a broad shallow sinus. One of our fossils exhibits a group of 15 tubes on a smooth space just under the reticulated portion of the sides, arranged as represented in De Koninck's figures of some examples from Vise, Belgium. (Monog. Prod, at Chon., Pl. IX, fig. 1b, c.)

In his later listb the form appears as P. semireticulatus var. antiquatus Martin. The spines which Shumard describes as being numerous just under the reticulated portion have not been observed by me, on account of the imperfect condition of my material.

---

bIdem, p. 389.

---

This variety is found not only in the white limestone, from which Shumard cites it, but also in the dark limestone. The only difference which my material shows between examples found at the two horizons is that the type from the white limestone is a little more finely ribbed than the other.

This form is by no means rare in the Capitan limestone, but all my material is more or less fragmentary. So extreme a range in size is indicated by the specimens that I suspect several varieties could be discriminated did not their imperfect condition render ineffectual an attempt to do so.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 3762); middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

PRODUCTUS MEXICANUS Shumard.

1858. Productus Mexicanus, Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 291 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

1859. Productus Mexicanus. Shumard, idem, p. 389.

White [Permian] limestone: Guadalupe Mountains.

?1877. Productus Mexicanus? White, U. S. Geog. Survey W. 100th Mer., Rept., vol. 4, p. 120, pl. 8, figs. 6a to 6c.

Carboniferous: Camp Cottonwood, old Mormon road, Lincoln County, Nev.; near Salt Lake, New Mexico.

?1883. Productus Mexicanus. Kayser, Richthofen's China, vol. 4, p. 182, pl. 28, figs. 7a-7b.

Upper Carboniferous: Lu Ping, China.

?1902. Productus Mexicanus. Tschernyschew Mém. Com. geol. St. Petersburg, vol. 16, No. 2, p. 264, pl. 52, fig. 10.

Schwagerina zone: Ural Mountains.

Shell of medium size, subrectangular, width greater than the length; dorsal valve elevated, strongly arched, marked with a broad, very slight mesial depression, which is scarcely developed into a sinus; sides rounded, falling abruptly to the margins, front very gently convex; beak small, pointed, convex, moderately prominent; surface ornamented with from 18 to 24 prominent, rounded, longitudinal ribs, their number somewhat increased by implantation or bifurcation. The ribs are separated from each other by spaces as wide as themselves, and both ribs and spaces are crossed at somewhat irregular intervals by rounded concentric folds, which give to the ribs at the points of crossing a handsome subnodulose character. The concentric folds are not as prominent as the ribs except on the sides, where, in one of the specimens before us, two or three of those nearest the border are developed into strong wrinkles. Ventral valve unknown.

Dimensions.—Length, 0.64; width, 0.70; height, 0.54. These proportions were taken from a young specimen on account of its being more perfect than the others. The collection contains fragments of full-grown shells, which, if perfect, would perhaps measure one-third greater.

White limestone of the Guadalupe Mountains.a

---

aTrans. Acad. Sci. St. Louis, vol. 1, 1856-1860, p. 291.

---

The foregoing are Shumard's description and remarks on this species, which, unfortunately, seems to be without representation in our collections. I know of no form having the large ribs and coarse wrinkles called for by Shumard, not to mention the specifications as to size and sinus.

White identifies this species from points in Nevada and New Mexico, and although his form agrees closely with Shumard's description in most respects except size (it being considerably smaller), it is probable that if representative specimens of P. mexicanus were at hand the two would not be found to be the same species.

The form most similar to P. mexicanus in the present collection is that designated Productus sp. a, which is from a different horizon. It seems to be closely related to P. mexicanus as identified by White, except that the ribs are coarser; but many characters of Productus sp. a are not well shown by the single imperfect specimen at hand. P. walcottianus resembles perhaps still more closely White's identification of Shumard's species. It has, however, stronger wrinkles over the visceral area, with more numerous spines and less regular and continuous ribs.

PRODUCTUS sp. c.

This division is made primarily for a single ventral valve from the Glass Mountains (station 3763). The posterior portion is enveloped in chert, so that the characters of this portion can not be made out. It is evidently nearer to Productus guadalupensis than to any of the other species recognized in this report, but differs in being somewhat larger, in having slightly coarser and stronger ribs, and in being less arched and inflated in the posterior portion. The ears appear to be large. The distribution and character of the spines has not been ascertained.

It differs from the large form of P. semireticulatus var. capitanensis in being smaller, and from both large and small forms in having a broader and shallower sinus and less prominent ribs. Its general shape is that of a species recently described by me as P. semireticulatus var. hermosanus. It is, however, slightly smaller, has a somewhat deeper sinus, and is more strongly arched in an anterior-posterior direction. The ribs are of about the same degree of coarseness, but it can not be told whether the posterior portion is crossed by wrinkles or not.

I have also provisionally referred here another ventral from the same locality, only the posterior portion of which has been preserved. The size and shape are such as might ally it with the foregoing. The ribs are peculiarly thin and high and separated by broadly rounded grooves wider than the ribs themselves. There are also fine concentric striæ, and concentric wrinkles which make nodes where they cross the ribs, and ridges between them, and are somewhat farther apart than the ribs themselves. The character of the costæ is not exactly that which would be expected to go with the larger specimen, in which the posterior portion is concealed, but not such as to preclude the association altogether.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

PRODUCTUS POPEI Shumard.

Pl. XX, figs. 9 to 11b.

1858. Productus Popei. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 290 (date of volume, 1860)

[Permian: New Mexico and Texas.]

1859. Productus Popei. Shumard, idem, p. 389, pl. 11, figs. 8a, 8b.

White [Permian] limestone: Guadalupe Mountains.

Shell of medium size, subquadrate, wider than long, greatest width at the cardinal border. Dorsal valve (receiving valve) gibbous, very strongly arched, somewhat inrolled, flattened convex near the beak; anterior prolongation of moderate length, forming a gentle curve from the visceral region to the front; sinus commencing near the beak, where it is very shallow, but it soon increases in depth, and becomes very profound on the anterior prolongation, so as to give this portion of the shell a very marked bilobed appearance; surface with from six to ten unequally rounded, coarse ribs on each side of the sinus, their number sometimes increased by division and implantation. These ribs are usually quite prominent and broad on the anterior prolongation, but on the posterior third of the shell they become obsolete, leaving a nearly smooth surface for some distance before the beak; sides falling abruptly to the margins, near which they are usually marked with a series of eight or nine rather strong tubes, which extend from the beak to the front. Besides these, most of the specimens exhibit a few smaller tubes sometimes scattered promiscuously over the surface, but generally ranging in oblique lines across the dorsum of the shell; beak small, pointed, slightly incurved, and passing a little beyond the cardinal margin. Ventral valve elliptico-subquadrate, gently concave or flattened on the visceral disk; its sides with a row of spines, which, with other surface ornaments, correspond to those of the opposite valve.

We dedicate this, one of the most beautiful species of the American Productus, in compliment to Capt. John Pope, of the United States Corps of Topographical Engineers, whose expedition has the honor of having first procured paleontobogical evidence of the existence of Permian strata in New Mexico and Texas.a

---

aTrans. Acad. Sci. St. Louis, vol. 5, 1856-1860, p. 290.

---

The foregoing is Shumard's original description of this species, quoted in full, in which I find it necessary to make few alterations. My specimens appear to have possessed less numerous spines than Shumard mentions, and in especial they lack any well-marked series of these developments on the sides. In fact the spines appear to be scattered, and are for the most part of small size, except the lateral ones. Shumard's statement that there are six to ten ribs on each side of the sinus is not very satisfactory, since the sinus in Productus can not be said to have any well-recognized limits. My specimens have about 20 ribs in all, those in the sinus being indistinct.

The description of the dorsal valve reads very much as if it were framed from observations of a flattened ventral, especially in the presence and arrangement of the spines. The only dorsal in our collection is similar to the ventral shells, but is flatter and with a less elevated and projecting beak. One specimen associated with the others of this species seems to mark a somewhat distinct variety, being narrower and with weaker ribs. It is represented by figs. 11, 11a, and 11b of Pl. XX.

On page 389 of the volume which contains the description quoted above, Shumard again remarks of this species:

A number of specimens of this species are in the collection, all of them from the white limestone of the Guadalupe Mountains. There are two distinct varieties; one having a remarkably deep sinus with 5 to 7 costæ on either side, and the other with a less profound sinus and from 8 to 13 costæ. The latter variety I at first regarded as a distinct species, but a more thorough examination of a number of specimens has led to the opinion that it should not be separated from the species above cited.

Shumard's figures are not so good but that a certain doubt remains,in my mind that this may not after all be P. popei, yet it is nearer than any other species in our collections. His figures show a narrower form, with more sloping sides. As to what the variety referred to in his later note may be, I am uncertain, unless it is the slightly larger but very similar species which I have described as P. popei var. opimus.

The shells here discriminated as P. popei and P. occidentalis are very similar to one another. Only minor differences exist, and larger collections may be expected to bridge them over. P. popei is smaller, with stronger and somewhat finer ribs. It is more strongly arched longitudinally, the posterior portion being flattened, with the rostral and umbonal areas depressed.

Horizon and locality.—"Dark limestone," Pine Spring (station 2930), and hill southwest of Guadalupe Point (station 2924?), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2957).

PRODUCTUS POPEI var. OPIMUS n. var.

Pl. XX, figs. 12 to 14a.

Associated with Productus popei in the "dark limestone" is a type that is related so closely to it that I am able to point out few characters save that of size by which they are distinguished; yet as the difference in size is rather marked, with few if any intermediate conditions, I am unwilling to refer them to the same species without some qualification. The following characters have been noted:

The shell is of medium size. Ventral valve transverse, the visceral portion flat, and the curvature strong and abrupt. Ears small, depressed, and probably slightly extended. Sinus rather deep and narrow, but undefined. Ribs moderately strong and coarse; about 21 on the type specimen. Spines few, large, chiefly on the lateral portion of the shell near the ears. In one specimen several occur in a row approximately defining the ears on each side.

Dorsal valve much like the ventral, but lower and with less elevated beak. The figured specimen has more regular and equal ribs than some of the others referred to the same species.

The shape of these shells is that which usually distinguishes dorsal valves of other species, and they might easily be mistaken for dorsals of P. occidentalis, which they resemble very closely. The chief difference seems to consist in the slightly greater prominence of the beak and slightly increased fineness of the ribs. The presence of spines on the convex sides of some of these shells, however, is conclusive evidence, in spite of their configuration, that they are ventral valves.

This form is readily distinguishable from P. occidentalis by its flattened shape and lower elevation. Some of the associated dorsal valves, however are with difficulty assigned to one or the other species, the chief difference being that they are much lower than P. occidentalis.

Practically the only character which distinguishes this form from P. popei is that it is distinctly larger. The configuration is practically the same. The ribs have almost the same prominence and degree of fineness, but as the shell is larger more of them are to be counted around the circumference.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PRODUCTUS INDENTATUS n. sp.

Pl. XX, figs. 15 to 16b.

Shell small, transverse. Ventral valve rather strongly arched, the visceral and anterior portions being more or less flattened, of equal length, and nearly at right angles to one another. Beak small, pointed, and slightly transgressing the hinge line. Ears large, flattened, projecting. Sinus narrow, shallow over the visceral portion, and deep anteriorly.

Shape of dorsal valve like that of the ventral, save that it is lower, with flatter visceral portion and less prominent beak.

Surface marked by radiating ribs of moderate strength and coarseness, or somewhat less, a few of which bifurcate about halfway forward from the beak. The bases of a few rather small spines occur on the ventral valve.

This form has about the size and something of the expression of the Marginiferas of the Mississippi Valley section; but such evidence as to its internal character as I have been able to gather would indicate that it is not a Marginifera. From the common Marginifera wabashensis or longispina of the Mississippi Valley it differs in having a deeper sinus, coarser and often stronger ribs, and fewer spines. I believe, however, that instead of being a Marginifera it is a Productus of the tye of P. popei et al.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PRODUCTUS TEXANUS n. sp.

Pl. XXI, figs. 25 to 26b.

Shell rather small. Ventral valve strongly arched longitudinally, subquadrate transversely, with a distinct but not very deep sinus and abruptly descending sides. Cardinal line extended. Ears rather large, convex.

Dorsal valve very transverse, beak apparently small and depressed. Ears large, projecting, defined by grooves, and upturned. Visceral portion flattened and marked by a narrow obscure sinus; which becomes deeper toward the anterior margin.

The surface is covered by strong, fine ribs, about five in 5 mm., which are obscure over the posterior portion. This area is nearly smooth, without either distinct ribs or distinct wrinkles. Toward the front of the ventral valve the ribs tend to become less pronounced and somewhat coarser, owing apparently to the fact that in some cases the older ribs unite or are replaced several by one. The spines are large and few in number, confined chiefly to the sides near the ears.

This shell somewhat recalls Marginifera wabashensis, but when compared with that species it proves to be larger, with coarser and stronger ribs. The dorsal valve is more transverse, with flatter visceral portion, larger ears, and deeper sinus. Indeed, the present species probably does not possess the characteristic structures of Marginifera.

It also resembles Productus indentatus, but is larger, less highly convex, with stronger ribs and shallower sinus. It stands rather close to Productus sp. d, with which it occurs in association at one locality. Its larger size and stronger, somewhat coarser ribs are its distinguishing characters, but if Productus sp. d has the reflexed border shown by a specimen supposed to belong to it, P. texanus is undoubtedly distinct.

Another closely related species is P. walcottianus. Considered on the basis of their type specimens, P. texanus is larger, with shallower sinus, larger and much less numerous spines, and nearly smooth instead of wrinkled and spinous visceral area. When only the anterior portion is examined the two species have a strong resemblance, but the posterior portions are somewhat different. While these differences rather strongly distinguish the typical specimens, it will be impossible to satisfactorily refer imperfect examples and casts.

In P. capacii D'Orbigny has described a shell related to the one under consideration. There is a slight difference in shape, the South American species contracting more rapidly above. The ribs appear to be finer and the surface more thickly covered with spines, but the two species are certainly closely related, and a comparison of specimens may prove them to be identical.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (stations 2903 and 2919).

PRODUCTUS sp. a.

Pl. XXI, figs. 24 to 24c.

This form is represented by a single small ventral valve having a width of about 16 mm. and a length of 13 mm. The sinus is rather strong toward the front, but indistinct on the posterior half of the shell. The ears are small, and ill defined. Beak rather inflated. Visceral area somewhat flattened. This portion of the shell is nearly smooth, though the more marginal areas are marked by coarse prominent ribs, about 13 or 14 in number. A few large spines are found on the sides and the ears.

This shell is about the same size as the Pennsylvanian species Marginifera wabashensis, but is easily distinguished by its stronger sinus and much coarser ribs. It more nearly resembles Productus indentatus of the present fauna, from which it is distinguished by its shallower sinus and considerably coarser ribs. Its general appearance is that of a diminutive example of popei.

Though from a considerably lower horizon, this is nearer to P. mexicanus Shumard than any other of the species recognized. It differs from the specimen identified and figured by White as P. mexicanus in being less strongly wrinkled on the visceral region and in having coarser ribs. All the characters, however, are not satisfactorily shown by the single example so far found.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (stations 2903 and 2919).

PRODUCTUS GUADALUPENSIS n. sp.

Pl. XXII, figs. 1 to 3a.

Shell of medium size, inflated, transverse. Ventral valve strongly arcuate longitudinally. Beak small, slightly projecting. Ears depressed, gently convex, extended. Sinus strong, narrow, and undefined. Surface marked by rather fine ribs, about five in the space of 5 mm., which are not very distinct and apparently die out altogether toward the anterior margin. They are indistinct over the posterior portion also, which is not crossed by well-marked concentric wrinkles, although this area has an obscurely nodose appearance. Spines comparatively large, few in number, and scattered over the more marginal portions of the surface. Dorsal valve not known.

This species has the general configuration of Productus multistriatus Meek, but it is more highly arched and has coarser ribs. Perhaps even more similar are the two South American species P. chandlessi and P. batesianus, both described by Derby. P. guadalupensis is somewhat smaller than P. chandlessi, with a stronger sinus, more projecting ears, and coarser ribs. It is somewhat larger than P. batesianus, with more projecting ears and finer ribs.

The typical examples of this species were found in the yellow sandstone of the Delaware Mountain formation at station 2919, and a single somewhat doubtful example was obtained from the same horizon at station 2931.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (stations 2919 and 2931).

PRODUCTUS GUADALUPENSIS var. COMANCHEANUS n. var.

Pl. XXXI, figs. 5 to 5b.

This form occurs in the Glass Mountains, and it so much resembles that which is found in the sandstones of the Delaware Mountain formation that I originally identified them as the same species. Their configuration is about the same, though the present form is perhaps a little more strongly arched, and it also has distinctly coarser ribs. The variety comancheanus, therefore, suggests Productus multistriatus much less than the typical one.

With but a single specimen of the present form and rather imperfect material representing the other, it is impossible to tell whether these differences are mere variations or are constant and associated with others. Consequently the present form has been introduced as a variety, and the matter left standing so for decision by fresh and more perfect material.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

PRODUCTUS OCCIDENTALIS Newberry.

Pl. XII, figs. 4 to 4c.

1858. Productus costatus. Marcou (non Sowerby), Geology of North America, p. 46, pl. 5, fig. 5.

Carboniferous: Cedar Creek, Mogollon Mountains; tributary of Gila River; sources of the Colorado Chiquito.

1861. Productus occidentalis. Newberry, Ives's Colorado Expl. Exped., Rept., p. 122, pl. 2, figs. 9, 10.

Upper Carboniferous (cherty limestone): Banks of Cascade River near junction of Great and Little Colorado.

This species is very abundant in the upper portion of the Aubrey, whose typical outcrops are found in northern Arizona, and it presents many variations, to not all of which the original definition of Newberry strictly applies. One of these varieties occurs in the white limestone of the Guadalupe Mountains, though but a single specimen, representing a ventral valve, has so far come to hand. Three rather imperfect specimens from the "dark limestone," which possibly do not belong to exactly the same varietal group, have also been collected. The more perfect example from the white limestone may be described as follows:

Shell of medium size, narrow, highly arched. Ears moderately large, flattened, depressed, more or less extended. Sinus strong, but broad and undefined. Ribs coarse and irregular, separated by rather wide intervals; not very strongly elevated, more or less nodulose and bifurcating. The spines for the most part are small and scattering, but on the sides near the ears a few large ones occur. The specimens from the "dark limestone" have ribs of nearly the same size, but more numerous and more closely arranged.

As previously remarked, this shell can be almost exactly matched by specimens from the Aubrey which I identify with Productus occidentalis. Of the different American forms which have been referred to P. costatus probably none approaches Sowerby's species more closely than does the one under discussion. Yet if P. costatus be restricted to the type with which the name was first associated none of these forms deserves to be referred to it. P. occidentalis has not been cited by name in paleontological papers since it was first described by Newberry, but it is probable that the form described and figured by Marcou as P. costatus is the same species. Marcou's figures appear to represent a dorsal valve, if one may judge by the apparently flattened posterior portion and absence of spines. He states that he had larger and better specimens, and his description appears to be based on a ventral valve. The resemblance of his figures to dorsal valves of P. occidentalis is striking, and the geologic occurrence which he cites is also favorable to their being the same species. Marcou's figures, however, represent the visceral portion of the valve as marked by concentric wrinkles, which is rarely the case with characteristic P. occidentalis.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

PRODUCTUS MEEKANUS n. sp.

Pl. XXX, figs. 13 and 13a.

Shell small. Ventral valve moderately convex. Beak small, inconspicuous. Ears small, quadrate, undefined. Sinus absent. Surface marked by concentric wrinkles, spines, and liræ. The wrinkles, which are strong on the sides and ears but fainter and somewhat interrupted elsewhere, cover practically all the surface. The spines are rather large, numerous, regularly arranged, and projecting strongly downward and forward, so that their bases have a "tear-drop" shape. The radiating liræ are fine and rather faint, tending to be interrupted by the spines, which form the most obvious feature of the surface ornamentation. Where preserved as internal molds in the yellow sandstone of the Delaware Mountain formation, specimens belonging to this species show little besides the spines, a few wrinkles near the ears, and the impressions of a few internal spinules about the margins.

P. meekanus is probably closely related to Productus signatus, but is constructed on a smaller and more delicate pattern. Its nearest representative in the Mississippi Valley is P. pertenuis Meek, although there can be no doubt that the two species are not identical. It is as if the spines, which are subordinate to the liræ and wrinkles in P. pertenuis, had in P. meekanus come to dominate them.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2931). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

PRODUCTUS SIGNATUS n. sp.

Pl. XXII, figs. 4 to 4b.

Shell of medium size, transverse. Ventral valve moderately arched, rather rapidly expanding. Dorsum broad and flattened, without a median sinus; falling off rapidly at the side. Ears apparently small, depressed, and rounded(?). Dorsal valve not known.

Surface covered with spines mounted on narrow elongate bases, which make up a sort of coarse, irregular, and discontinuous ribbing. There are also moderately fine, somewhat indistinct, transverse wrinkles covering nearly the whole shell. On the internal mold, the preservation in which this fossil occurs, the surface between the spine bases is marked by fine, discontinuous, longitudinal liræ, especially on the anterior parts, and by pits produced by numerous small spinules on the inner surface of the shell. These fine liræ, like the spinules, may be entirely internal structures, but they may also be and more probably are the expression on the inside of fine intermediate liræ, like those on the surface of Productus meekanus.

An external mold of a dorsal valve is marked by moderately fine concentric wrinkles, fine, interrupted radiating liræ, and numerous slender spines, a feature not very common for the dorsal valve even when they occur on the opposite one.

This species is allied to P. opuntia Waagen, but can be distinguished by the more elongated character of the spine bases, from the middle of which the spines arise.

It is closely related to P. meekanus, but the sculpture, while of the same general character, is on a much larger scale.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2931).

PRODUCTUS SIGNATUS var.

Associated with the form which I have described as Productus signatus is another very similar to it, which, while considerably smaller in size, has the surface ornamentation of the same character and on about the same scale. The internal spinules are much more numerous on the anterior and lateral portions of the shell.

The width of the specimen for which this division is made is about 25 mm. and the length from the anterior margin to the point of greatest convexity 22 mm. The size, therefore, is considerably smaller than that of P. signatus, and about that of P. meekanus, but the surface ornamentation, in the wider spaces at which the spines stand, rather resembles the former species.

Too little is known of these forms to permit a conclusion as to their actual affinities.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2931).

PRODUCTUS LATIDORSATUS n. sp.

Pl. XI, figs. 11 to 13b.

Shell rather small, thin, transverse. Ventral valve moderately convex, the anterolateral curvature being stronger than the transverse. Beak small, incurved, not projecting far beyond the hinge line. Ears small, depressed, somewhat arched, probably slightly projecting. Sinus shallow, ill defined.

Dorsal valve moderately convex, with flattened visceral region. Beak inconspicuous. Ears small, depressed, somewhat arched, and slightly projecting.

Surface nearly smooth. Traces of ribs can be seen on most specimens, but they are usually faint. Equally or somewhat more indistinct concentric wrinkles also are found, and rather strong growth lines. Spines are numerous and small over the posterior portion of the shell, rather scarce and larger over the front, lateral portions, and ears. The ribs in many cases, especially where at all well marked, are more or less discontinuous, and originate severally at the bases of the larger spines.

This species is related to P. wallacianus Derby, but it is distinguished by having a more or less distinct sinus and much less numerous spines. It is also allied to P. subhorridus Meek, but before comparing the Guadalupian form with that species it will be necessary to define the latter a little more closely than Meek has done.

Meek apparently figures two specimens of P. subhorridus.a One specimen was probably the original of figs. 3 and 3b, and another of fig. 3a. The description of plates seems to indicate that 3 and 3b represent different specimens, but the undoubted original of 3b corresponds exactly to fig. 3, whereas no other specimen in the type lot is at all like it. But the surface of the specimen which undoubtedly stood as the original of 3b is unlike the illustrations of either 3b or 3. It is in fact largely exfoliated, and the spines represented on both figures are not found on the specimen. I have been able to find traces of less than ten on the entire surface, and it is safe to say that they are much less numerous than represented. Furthermore, on parts of the shell, and especially on a small area on the sides where exfoliation appears not to have touched, the surface is marked by very fine, regular, equal, radiating liræ. I can not be sure, without other specimens, of the validity of these observations as applying to the real surface ornamentation, but these differences, taken with the more slender form and deeper sinus, indicate rather strongly that the original of figs. 3 and 3b does not belong to the same species as that of fig. 3a. To the latter it is proposed to restrict the name P. subhorridus, because to it belong most of the type lot of specimens, and because the name subhorridus and the description clearly apply to this type. Perhaps the only character in Meek's description which is taken from the original of figs. 3 and 3b is the strong sinus which is ascribed to P. subhorridus, a character which is far from distinct in shells having the abundantly spinose surface of fig. 3a.

---

aU. S. Geol. Explor. 40th Par., vol. 4, 1877, p. 75, pl. 7, figs. 3, 3a, and 3b.

---

From P. subhorridus thus restricted P. latidorsatus differs in being shorter and broader, with much fewer spines.

This species is plentiful in the white limestone, though apparently it was not represented in Shumard's collections. No specimens have been found belonging to it in the "dark limestone," but to the same species has been referred, with some doubt, an imperfect specimen from the Delaware Mountain formation. This example is similar to immature stages of latidorsatus, but has rounded cardinal angles and appears to be without sinus, ribs, or spines. The preservation, however, is so imperfect that these characters might be largely lost.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); Delaware Mountain formation, Guadalupe Point (station 2903), Guadalupe Mountains, Texas.

PRODUCTUS LATIDORSATUS var.

A few specimens from the black limestone of the Guadalupian section are related to Productus subhorridus, P. subhorridus var. rugatulus, and P. latidorsatus, but my material is too insufficient and imperfect to permit me satisfactorily to ascertain in what degree. These examples can not be precisely identified with any one of the three types mentioned. They possess the size and general expression of P. subhorridus var. rugatulus, but have a fainter sinus or none at all, lack the posterior wrinkles, and have numerous small spines over the lateral and posterior portions, with a few large ones near the ears. So far as can now be told they agree in every essential particular with P. latidorsatus except that they are considerably smaller. On this account and because of the wide difference in stratigraphic occurrence I feel indisposed to unite them directly with P. latidorsatus, and it is probable that more abundant material will show that they are a distinct species.

In the Aubrey beds of Utah, Arizona, and New Mexico is found a series of closely connected varieties which probably embraces P. subhorridus Meek, together with the form figured by Newberry as P. costatoides Swallow, and of whose members some more or less closely resemble P. latidorsatus, P. latidorsatus var., P. subhorridus var. rugatulus, and P. walcottianus. A rather hurried review of this material indicates that its range of variation approaches though it does not include P. latidorsatus and P. walcottianus. Some types come very close to P. subhorridus var. rugatulus, but few if any of the Aubrey forms show the wrinkled posterior surface of that species.

While this almost uninterrupted series embraces P. subhorridus also, I would be unwilling to subsume the Guadalupian form under Meek's species as strictly identical with his types. The more careful discussion of the Aubrey forms and their relation to the species in hand is reserved for another occasion.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2920).

PRODUCTUS sp. e.

Pl. XXII, figs. 5 and 5a.

This division is established for a dorsal valve which I am unable to refer to any of the species recognized in this report. It has a subquadrate shape, with the hinge a little longer than the width below. The front and sides are somewhat rectilinear, but unite in a strong curvature. The ears are small, depressed, defined by grooves, slightly upturned. The beak is small, prominent; the sinus moderately strong. The surface is marked by a number of small prominences (seen as such on molds) and small spines (seen as holes). No ribs are visible, and only extremely faint traces of concentric markings.

These notes being based on a mold of the exterior, the characters as described would in reality of course all be reversed.

This specimen was associated with Productus signatus and P. signatus var., but from its size, surface ornamentation, and sinus it can hardly belong to either of them. It was also associated with specimens some of which have been identified with P. subhorridus var. rugatulus and P. walcottianus. Typical dorsal valves of the former are marked by fine but very distinct wrinkles, and are without spines on the exterior, so that the specimen under consideration can hardly be referred to the same species. The dorsal valve of typical P. walcottianus is not known, but it could hardly fail to show faint indications of ribs. Dorsal valves that seem to belong with the ventrals from station 2931 that have been referred to P. walcottianus differ from the specimen before me in having numerous concentric wrinkles, together with finer and much more numerous spinules. If they properly belong to P. walcottianus, as I somewhat doubt, this certainly does not. It is perhaps nearest to P. latidorsatus, but it is not probable that the dorsal valve of that species is provided with spines.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2931).

PRODUCTUS SUBHORRIDUS var. RUGATULUS n. var.

Pl. XXX, figs. 11 to 12c.

Shell small, transversely subquadrate. Ventral valve moderately convex. Beak rather prominent and projecting. Ears small, depressed, inconspicuous, and undefined; not extended. Sinus usually present, but variable in strength, sometimes well marked.

Dorsal valve shallow, gently concave over the visceral area, more strongly bent toward the margins. Ears small and undefined.

Surface of the ventral valve marked by numerous concentric wrinkles, which are sometimes moderately strong over the visceral region, but are evanescent toward the anterior margin. Numerous strong growth lines are also present, more or less completely graduating into wrinkles. There are no well-defined radiating striæ, but coarse, indistinct ribs originate at the bases of the spines, and maintain themselves some distance. The spines are rather large, long, and scattered. The surface of the dorsal valve is marked by numerous fine wrinkles and by a number of large shallow pits answering to the spines on the surface of the other valve. On the interior of the ventral valve the line along which the ears would be separated from the body of the shell is marked by a crenulated ridge, as in Marginifera; but the dorsal valve does not possess the beveled submarginal band characteristic of that genus.

This species has been obtained at several points, but the typical examples were collected in the Glass Mountains (station 3763), where it is especially abundant and perfect. I have referred to it, though with some hesitation, a few specimens from the yellow calcareous sandstones of the Delaware Mountain formation (station 2930), where the fossils occur as molds. The abundant form at this locality appears to belong to Productus walcottianus, and I have not been able to satisfy myself, in the imperfect condition of the shells placed with this species, that they are not extreme variations of the other type, whose characters have been obscured and modified by their preservation. They have about the size and configuration of P. subhorridus var. rugatulus, and the spines and coarse indistinct ribs of that species. There is, however, little or no indication of a sinus or of concentric wrinkles on the posterior portion. The latter feature, as I have suggested in the case of the associated fossils referred to P. walcottianus, which also has a wrinkled posterior portion, is doubtless partly obscured by their preservation as molds. It can not be seen on the inner side of typical examples of P. subhorridus var. rugatulus. It is evident, however, that variations are indicated by the material of which I have not been able to take cognizance on account of poor preservation, and that there will be difficulty in the case of ill-preserved or imperfect specimens in discriminating between P. walcottianus, P. subhorridus var. rugatulus, and possibly P. latidorsatus.

This species is similar in many ways to certain of the forms belonging to Marginifera, but lacks some of the internal structures characterizing them. A form almost identical is found in a collection from Lake Titicaca, Bolivia. It resembles also P. latidorsatus, but can be distinguished by being very much smaller, with sinus slightly more marked, relatively larger and more numerous spines, coarser ribs, and stronger wrinkles. Certain small and narrow varieties of P. subhorridus, which runs through numerous variations, resemble this form very closely, but it can usually be distinguished by its more strongly wrinkled posterior surface. From typical P. subhorridus it is distinguished by being much smaller, by having the posterior portion crossed by fine wrinkles, and by having much fewer spines.

It would be desirable to compare this form with P. costatoides if it were possible to do so. Swallow describes but does not figure his species, so that I am placed at a disadvantage in attempting to compare the two forms. His description indicates a species of the general character of Marginifera wabashensis, M. splendens, and their allies, and, indeed, his comparisons are with them. In spite of the fact that Swallow remarks that P. costatoides has not the flat band of M. splendens, I suspect that it belongs to the same group. In many of the Marginiferæ this is not a conspicuous character. The rugæ, striæ, spines, and pits described by Swallow are, however, somewhat different from the ordinary type of M. splendens and M. wabashensis, and if accurately set down indicate a species distinct from them. The form must at the same time be a rare one, as it has never since been recognized, nor do I recall having seen it among the very numerous representatives of that and related groups which have passed under my observation. Newberry identifies and figures P. costatoides from the Grand Canyon country, but it is doubtful if his identification should carry weight of absolute authority. The form figured by him is merely one phase of a very variable series whose members are common in the Aubrey of Arizona, Utah, and New Mexico, and of which some aspects will probably include the type which Meek described as P. subhorridus.

The variety of the latter species here under discussion resembles these shells very closely, but seems to be distinguished by the fine wrinkles by which the visceral region of both valves is covered. Its affinities with the typical P. costatoides can not be definitely determined until specimens of the latter are obtained for comparison. The two forms seem to be closely related, but differences as well as resemblances are indicated by Swallow's description. P. subhorridus var. rugatulus is apparently a somewhat larger form than P. costatoides. The visceral region of the ventral valve is not flattened; the sinus, which Swallow describes as deep and broad, is less well marked; the ears are smaller; and the costæ are fainter. In view of the fact that the western form probably occurs at a higher horizon than P. costatoides, which is not even referred to the Permian, but to the upper coal measures, and of the differences just noted, there is a reasonable probability that if it is ever ascertained just what form Swallow had in hand—which can not be satisfactorily told from his description alone—it will be found to be distinct from that here called P. subhorridus var. rugatulus. At all events, until P. costatoides is better known, it seems to me that it will be advantageous to leave it in abeyance, so that the descriptive side of the discussion may be on as stable a footing as possible.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2931?). Delaware Mountain formation, Comanche Canyon, Glass Mountains (station 3763), and mountains northwest of Marathon (station 3840), Texas.

PRODUCTUS WALCOTTIANUS n. sp.

Pl. XXI, figs. 27 to 28b.

Ventral valve small, narrow, high, very convex, the summit of the curvature being about two-thirds the distance back from the anterior margin. Anterior and posterior slopes flattened. Dorsum broad and planate, with a rather strong sinus. The sides descend rapidly. The ears are probably rather small and not much extended. Beak somewhat large and tumid, though passing but little beyond the hinge line. The visceral portion is strongly but irregularly rugose and thickly covered with rather large spine bases, giving it a pustulose appearance. The anterior half of the shell is marked by rather strong and coarse but not very regular ribs. The spines are especially abundant on the visceral and lateral portions of the shell, but are scattered over the entire surface, becoming less and less numerous toward the anterior margin. Dorsal valve unknown.

This species is primarily based on a unique but well-preserved ventral valve found at station 2903, in the Delaware Mountain formation. At station 2931, also in the Delaware Mountain formation, though at a different horizon, occurs, in considerable abundance, a similar form, which I have with some hesitation referred to the same species. Satisfactory comparison of the two is difficult, because in the one case the specimens appear as internal molds, and in the other they retain the shell. Certain departures from the type specimen can be made out. In some instances the outline contracts more rapidly above, instead of having nearly parallel sides. The surface of some of the molds is also nearly smooth, indicating that the ribs were not so strong as in the type. Evidence of the same thing appears on molds of the outside, and numerous small spines are seen to be mounted on the ribs.

Dorsal valves from this station seen as molds show a low elevation and are either almost regularly curved or flattened with a stronger marginal bend. Sinus rather faint. Surface marked by more or less strong concentric wrinkles, with numerous small depressions leading down to the cavities left by spines, which are arranged in concentric rows.

This species has about the size and configuration of P. subhorridus var. rugatulus, but the ornamentation is different. The concentric wrinkles are coarser and stronger and the spines more numerous. While it is improbable that there will ever be much difficulty in discriminating the two forms when fairly good specimens are in question, I have been troubled to dispose of the fossils from the sandstones of station 2931 in a manner satisfactory to myself. These are mostly internal molds, and besides imperfectly representing the surface characters of the original shell seem to indicate a certain amount of modification in the characters which distinguish the two forms. I have referred to P. subhorridus var. rugatulus, though with considerable hesitation, a few specimens from this station having a nearly smooth surface, with a few coarse plications.

This species shows many points of resemblance with White's identification of P. mexicanus Shumard. It differs in having larger and more numerous spines, discontinuous ribs, and stronger wrinkles over the visceral area. Even therefore if White's identification proves correct, as seems rather doubtful, P. walcottianus can hardly be referred to the same species.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (stations 2903 and 2931). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969?).

PRODUCTUS? PILEOLUS Shumard.

Pl. XII, figs. 8 to 15a; Pl. XXIX, figs. 5 to 7a.

1858. Productus pileolus. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 291 (date of volume, 1860).

White Permian limestone: Guadalupe Mountains.

1859. Productus pileolus. Shumard, idem, p. 389.

White [Permian] limestone: Guadalupe Mountains.

Shell small, strongly arched, gibbous, outline approaching to subquadrate, length greater than the width, widest at the cardinal margin. Dorsal valve gibbous, without mesial sinus, sides and front rounded, terminating below in a projecting band or rim, which is rounded and extends to the cardinal edge; umbo prominent, somewhat flattened anterior to the beak, slopes falling rather abruptly to the ears; beak prominent, rounded, strongly incurved, passing beyond the cardinal border; ears triangular, of medium size, incurved at the cardinal edge, convex in the middle and depressed at their junction with the umbones; surface of visceral region marked with several slightly elevated, concentric folds, which are most prominent on the sides and are continued on the ears, where they are directed backward and become obsolete before reaching the cardinal edge; anterior prolongation smooth or marked with very obscure concentric folds.

The collection contains but one specimen of this little species and this is partially deprived of its test.

Dimensions.—Length, 0.36; width, 0.32; height, 0.24.

Occurs in the white Permian limestone of the Guadalupe Mountains.a

---

aTrans. Acad. Sci., St. Louis, vol. 1, 1856-1860, p. 291.

---

The foregoing is Shumard's original description of Productus pileolus. The shells which on revision I have placed with this species I originally described as new because of certain departures which they show from the characters as defined above. Imprimis, I think that we must conclude that through some oversight Shumard wrote "dorsal" for "ventral" valve in his description, for that the single specimen on which he based the species was really a ventral is pretty clearly shown by its convexity (given in figures as nearly equal to the width), by the prominence of the umbo, and by the fact that the beak is described as passing beyond the hinge line. The determination of this fact is of some importance both in the matter of specific identification and in the interpretation of the structure, which Shumard describes as a projecting band or rim. If his specimen was a ventral valve the nature of this feature is somewhat obscure. It suggests a configuration resembling Productus limbatus or Productus sp. d. In the dorsal valve one would feel warranted in identifying it as the characteristic structure of Marginifera.

On the ventral valve of this species, in some cases at least, is found a feature which is rather unusual and of which I do not understand the significance. After flattening out on the ears as in all species of the genus, on reaching a certain stage, which must mark the limit of width of the hinge line, the shell again assumes a downward direction of growth, a change which of course is imperceptible at the front but becomes more and more obvious toward the cardinal line. The specimen represented by fig. 15 has this character well developed and other specimens have more or less distinct traces of it which have not been shown in my figures. It is probably this structure which Shumard refers to in the passage just discussed.

The fact that my specimens did not altogether agree with Shumard's description has already been remarked. That they nevertheless belong to the same species is rendered probable by the following considerations. They represent a not uncommon species in the Capitan fauna, and it is probable that Shumard's collection would contain one or more specimens. Again, they agree in a great many points with Productus pileolus; and, finally, some material probably belonging to the same species which was received subsequent to my original studies bridges over some of the differences at first noticed. My specimens differ considerably in proportion, but as a rule they are distinctly elongate. Shumard described the shape as widest at the hinge, but much of my material from the Capitan limestone appears to contract toward the cardinal line, and this is especially noticeable in dorsal valves. Considerable variation, however, is manifested in this particular, and in the recently obtained fossils from the southern Delawares (station 2969) the width of the hinge is a noticeable feature. Yet I can not consent to separate these later examples from those from the Capitan limestone.

Shumard's description of the projecting band or rim is still difficult for me to comprehend, though some of my specimens show suggestions of a marginal prolongation, which may be what he observed. The concentric folds which Shumard mentions as a feature of this species are prominent on one or two specimens, but as a rule are obscure or absent. The surface otherwise appears to have been smooth, without a sinus, without ribs, and practically without spines, though in several cases the base of a large spine was seen on one side about halfway toward the front.

Productus pileolus is very similar to the shell from the Trogkofelschichten which Schellwien described as P. curvirostris, but it differs in configuration as well as in having less numerous spines. In any event the Guadalupian species has long priority of publication. Indeed, it is not certain that the two species belong to the same genus, for Schellwien refers P. curvirostris to Productus, while the species under consideration contains traces of submarginal ridges, which may make it necessary to refer it to Marginifera. In that event it would probably belong in a group the same or related to that which Waagen calls the group of Marginifera spinosicostata. It is very different from our Pennsylvanian Marginiferas of the Mississippi Valley.

Shumard's species came from the Capitan limestone of the Guadalupe Mountains, and the specimens which I especially place in the same species came from the same locality and horizon. A single example from Shumard's "dark limestone" has also been placed here. It is not very perfect, but is a narrow form, more highly arched, and with a more prominent beak and umbo than those from the Capitan. If it should prove to be typical of a series of forms at this horizon, I would be disposed to regard it as a distinct variety or even species. This specimen has very nearly the configuration of P. curvirostris, but has fewer spines and no traces of ribs.

Some silicified specimens obtained south of the Guadalupe Mountains, at a horizon supposed to be about equivalent to the "dark limestone," are unfortunately mostly dorsal valves, the characters of which are very suggestive of this species. The only difference which I can at present point out is a relatively wider hinge line, the very point in which typical Productus? pileolus is reported as differing from my specimens. It has not seemed expedient to distinguish these older shells from those from the Capitan formation, though better material may render such a course necessary.

Horizon and locality.—Middle of the Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930?), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

PRODUCTUS PINNIFORMIS n. sp.

Pl. XII, figs. 5 to 5b.

Shell small, spatulate. Sides more or less straight, and diverging at an angle of somewhat less than 90°. Lateral outline merging with the anterior in a broad, full curve. Convexity moderate; greatest near the beak, which is curved downward somewhat abruptly. Length 20 mm.; greatest width 18 or 19 mm.

The entire surface is marked by rather strong but irregular concentric wrinkles and by very fine radiating ribs, of which from 9 to 11 occur in the space of 2 mm. They are wavy, and increase somewhat irregularly by implantation, so that the number to be counted within a given linear distance is not everywhere the same. They are rounded and separated by rounded intervals of about their own thickness. No spines have been observed.

This species is most nearly allied to P. campressus Waagen, which it resembles both in configuration and sculpture. It is, however, a much smaller form and considerably less elongate and apparently more finely ribbed. In view of these differences, of the fact that I have no specimens of P. compressus to furnish the final evidence, and of the fact that the American fauna contains so few identical species, though many similar to those in the Salt Range, it has seemed best to assign to the Guadalupian form a new name. I have but a single specimen of this species—apparently a ventral valve—and it is, moreover, rather imperfect; but it forms a rather interesting feature of the Guadalupian fauna.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

PRODUCTUS LIMBATUS n. sp.

Pl. XX, figs. 17 to 18.

Shell of medium size, transverse, shortest at the hinge line.

Ventral valve of rather low convexity, flattened over the visceral portion, suddenly curved around its margin. Beak low, pointed, not projecting. The ears are depressed, flat, and extended as a sort of band part way down the sides.

Dorsal valve similar to the ventral, but with flatter visceral area and less distinct beak.

Surface nearly smooth. It is marked marginally by faint, moderately fine ribs, which extend to a greater or less distance up onto the visceral area before dying out, but are not found on the ears. Except for faint concentric striæ the posterior portion of the shell in both valves is practically smooth. A few rather small spines spring from the lateral portions of the ventral valve.

Productus limbatus resembles P. pileolus Shumard, but as he describes that species as being strongly arched and widest at the hinge line, it seems unlikely that the two forms can be the same, especially as P. pileolus is much smaller. The fact that P. limbatus contracts strongly at the hinge line renders it almost unique among the species which have come under my observation.

This is a rare form, only two examples having so far been obtained, and it appears to be confined to Shumard's "dark limestone."

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PRODUCTUS sp. d.

Pl. XX, figs. 19 to 21a; Pl. XXIX, figs. 4 to 4b.

This is another of the Producti having a flattened visceral portion surrounded by an abrupt, strong curvature amounting in some cases almost to geniculation, a type of shell rather distinctive of the Guadalupian fauna; but, unlike many of the species, the present one does not possess a deep, narrow sinus. It appears to be new, but as my material is unsatisfactory no distinctive name is proposed for it. The following description has been drawn up:

Shell rather small, subquadrate, transverse. The ventral valve is not very convex, the visceral portion being somewhat flattened. The front and lateral margins are low, and likewise flattened, the curvature strongest about the edges of the visceral area. The beak is moderately elevated and slightly projecting (?). The ears are small, depressed, and quadrate. A shallow undefined sinus begins toward the front of the visceral area.

The dorsal valve has much the character of the ventral, except that the visceral area is flatter and the beak scarcely at all elevated.

The surface over the visceral region is almost smooth, without either distinct ribs or concentric wrinkles. The front and sides are marked by fine, somewhat faint, ribs, which begin toward the margin of the visceral area. They number about six or eight in 5 mm. The ventral valve for the most part appears to be without spines, though a few of large size are found on the ears near the cardinal line.

The ventral valve of this form is much like the dorsal valve of Marginifera wabashensis, and of course considerably different from ventral valves of that species. The dorsal valve is correspondingly different from the dorsal valve of M. wabashensis. In addition to the difference in configuration, the ribs as a rule are coarser and stronger and the spines are less numerous. While there is a certain superficial resemblance to Norwood and Pratten's species, I have not been able to find the distinctive submarginal ridges, so that it probably can not be referred to Marginifera at all. It may belong to the group of Productus popei, and it resembles P. indentatus, from which it is distinguished by being larger and more finely ribbed and by having a shallower sinus.

I have also referred to this species an interesting but fragmentary specimen from a locality considerably south of the typical Guadalupian section (station 2969). It is represented by fig. 4 of Pl. XXIX. Although so imperfect, this example is clearly a ventral valve, for the convex side is marked by fine, indistinct ribs, still less obvious on the concave surface, which is, moreover, provided with little spinous projections, such as frequently cover portions of the inside of Productus shells. This example possesses the rare feature of having, as represented in the figures, a strongly recurved margin. This rim looks at first like part of another valve closed upon the less fragmentary portion, but not only is it continuous with the latter, but it shows spinules on its convex surface, and in view of the fact that the rest of the specimen is a ventral valve, it could have no other interpretation than that adopted here, unless the rim were supposed to be part of another individual. This, however, is entirely out of the question. If this specimen really belongs to Productus sp. d, there can be little doubt that the latter is an undescribed form and one which on account of this singular feature can not be assembled in the group of P. popei. Slight indications are found in other examples that they also have a reflexed margin.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Genus STROPHALOSIA King.

I can not but feel that the genera Strophalosia and Aulosteges are at present on an unsatisfactory footing, although I have nothing to add to the situation likely to improve it. There seem to be in the Carboniferous faunas two types of shells which have a productoid expression and yet possess a more or less elevated area in both valves. One of these consists for the most part of small forms with low, imperfectly developed area, which are either cemented to some larger organism or marked by a distinct scar of attachment. The other is larger, with a high and well-developed area in many species, and without evidence of attachment. This form is apt to have a long and pointed beak, while the other, by reason of its lower area and its habit of cementation by this part of the shell, is apt to have the posterior extremity rounded or truncated. The internal structures of these forms are often impossible to determine, and where ascertained the different species seem to show great variability. Hall and Clarke appear disposed to regard the habit of attachment peculiar to one group of forms as the most important or at all events the most practical differentiating character, and they use the term Strophalosia for the cemented and Aulosteges for the uncemented forms. This seems to conform with the common usage in this country, where, with the exception of Shumard's little-known Guadalupian species Aulosteges guadalupensis, the type called Aulosteges does not occur.

When, however, I consider the typical species with which the names Aulosteges and Strophalosia must be associated, I confess to feeling some doubt as to whether they are not the same thing and are not, coincidentally, distinct from the shells which we have been accustomed to place with Strophalosia. This opinion is almost exclusively formed on the literature, for these types are either entirely lacking to our faunas or rare and imperfect. Figures of Strophalosia excavata, the typical species of Strophalosia, and of Aulosteges wangenheimi, the typical species of Aulosteges, certainly show the same general character and expression, one in a measure distinct from the small cemented American shells, which, for the most part of an earlier geologic period, it has been the practice to place under Strophalosia. Nor does it seem to me that the structural differences which have thus far been pointed out between the Permian Aulosteges and Strophalosias are sufficiently constant and well marked adequately to distinguish the two genera. It would rather appear to me, therefore, from an imperfect acquaintance with the subject that the small shells commonly assigned to Strophalosia may prove a distinct and possibly a new genus, and that typical Aulosteges and Strophalosia are really the same, Strophalosia being the older name. I have not, however, had sufficient confidence in this conclusion to act on it and have continued to follow the authorities in using Strophalosia for the small attached forms, although it involves what seems an inconsistency, since necessarily included with them are the apparently different typical Strophalosias of the European Dyas and Permian, which are at least very closely similar to the typical Aulosteges.

In the Guadalupian these two groups are fairly distinct in point of configuration, although the determination of internal or structural characters has not been possible. The small attached shells which I have called Strophalosia hystricula n. sp. and S. cornelliana Derby represent one type and the imperfectly known Aulosteges guadalupensis another. A. medlicottianus var. americanus, A. magnicostatus, Aulosteges sp. a, and Aulosteges sp. b are different from either of these forms in sculpture and configuration, but because of intrinsic character as well as of relation to foreign species it has seemed best to place them also with Aulosteges.

STROPHALOSIA HYSTRICULA n. sp.

Pl. XXX, figs. 14 and 14a.

Shell rather small, subcircular. Convexity moderate. Beak rather elevated. Hinge line shorter than the shell below. Ears small, undefined. Area low. Delthyrium narrow. Dorsal valve not known.

Surface covered by large tubular spines, most of which are straight and bent forward so as to be tangent to the surface. A few distributed over the surface, especially on the ears, are erect and sinuous. There are also a few irregular non-persistent concentric wrinkles.

This species finds its nearest ally probably in S. cornelliana Derby, from which it can be distinguished by having a more distinct and elevated beak and larger and less numerous spines, fewer of which are reclining. It also appears to lack the lamellæ described by Derby.

Strophalosia hystricula likewise resembles the form figured by Hall and Clarke as Strophalosia spondyliformis White and St. John. The real S. spondyliformis is a tiny shell, and so inadequately described and figured that comparisons can not be undertaken without specimens on which to base further observations. It is by no means certain that Hall and Clarke's identification in the work above cited is correct.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

STROPHALOSIA CORNELLIANA Derby.

Pl. IV, figs. 4 to 5.

1874. Strophalosia Cornelliana. Derby, Bull. Cornell Univ., vol. 1, p. 45, pl. 3, figs. 28, 30, 32, 33, 35-38; pl. 4, fig. 5; pl. 8, fig. 17; pl. 9, figs. 10, 11.

"Coal Measures:" Bomjardin, Brazil.

1892. Strophalosia Cornelliana. Hall and Clarke, Nat. Hist. New York, Pal., vol. 8, pt. 1, pl. 15B, figs. 36, 37.

"Coal Measures:" Bomjardin, Brazil.

Only a single imperfect example has so far been obtained, the origin of which was in the white limestone of the Capitan formation southwest of Guadalupe Peak. It has no marked characters, except the surface, which is thickly covered with small spines. Many of these lie flat and are nearly tangential. There are also concentric wrinkles more or less concealed by other superficial markings.

So far as its characters are demonstrated by the specimen before me, this form is so closely related to S. cornelliana, a figure of which is introduced for comparison, that it seems necessary to refer it to Derby's species.

Horizon and locality.—Base of Capitan formation, hill southwest of Guadalupe Point, Guadalupe Mountains, Texas (station 2906).

STROPHALOSIA sp.

Under this title I am subsuming two small and somewhat imperfect shells whose generic relations are evidently with Strophalosia cornelliana. They appear to be related to it specifically also. One of these specimens is copiously covered with large spines, all of which are erect, instead of being, as in S. cornelliana, in part appressed. The surface between the spines is very irregular, marked both by fine, discontinuous, concentric wrinkles and fine discontinuous ribs leading down from the spine bases. The other specimen has similar erect spines, but only a few of them, most of the surface having an uneven warty appearance. This may not be the normal sculpture, however, so provisionally the two forms have been left under the same caption.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2957 and 2969).

Genus AULOSTEGES Helmersen.

As noted in connection with the genus Strophalosia, I am using that term and Aulosteges in a different sense from that which, from a necessarily imperfect knowledge, really appears to me proper. It appears to me in fact that typical Strophalosia and Aulosteges are the same, and that Strophalosia, which is the older name, should be employed for the present forms to the exclusion of those which, as species of Strophalosia, have been recognized in the Guadalupian fauna. The latter seem to be congeneric with species of much older geologic periods—for example, with some which occur in the Mississippian of the Mississippi Valley; but the present group, here called Aulosteges and believed to include the typical Strophalosias of the European Permian, appears to be restricted to the latest faunas of the Paleozoic.

There can be little doubt that Shumard was correct in referring to this genus Aulosteges guadalupensis. Four other species contained in our recent collections with scarcely less certainty belong here. These species, five in all, are the only ones at present known from either of the Americans. As the genus is regarded as a typical Permian one, its presence in the Guadalupian fauna is interesting from a stratigraphic point of view, especially since it occurs in the earliest as well as the latest beds of the series; but the material thus far obtained is too scanty and too poorly preserved to add anything in the way of biology or structure.

The five species found in the Guadalupian fauna represent four rather distinctly marked types. To one of these belongs only the form which Shumard described so many years ago, Aulosteges guadalupensis. The second comprises A. medlicottianus var. americanus and Aulosteges sp. a. The third is represented only by the imperfectly known Aulosteges sp. b, and the fourth only by A. magnicostatus.

AULOSTEGES GUADALUPENSIS Shumard.

Pl. XX, figs. 22 and 22a.

1858. Aulosteges Guadalupensis. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 292 (date of volume 1860).

White [Permian] limestone: Guadalupe Mountains, New Mexico and Texas.

1859. Strophalosia (Aulosteges) Guadalupensis. Shumard, idem, p. 390, pl. 11, figs. 5a, 5b.

White [Permian] limestone: Guadalupe Mountains.

1890. Strophalosia? Guadalupensis. Beecher, Am. Jour. Sci., 3d ser., vol. 40, p. 241.

Ventral valve large, outline subelliptical, gibbous, flattened convex at the umbo, enlarging rapidly from beak to front and forming a pretty regular curve in the same direction; greatest width about the middle of the valve; lateral margins rounded, front slightly sinuate; a broad, shallow sinus commences some distance in advance of the beak and continues to the front in one of the specimens, and in the other the sinus is somewhat profound and narrow on the umbonal region and becomes shallow toward the front; beak elongated, flattened, straight or slightly curved upward at the extremity, which is pointed; area triangular, very much elevated; lateral edges sharp and strongly defined. Surface marked with numerous slightly prominent, radiating, interrupted ribs, crossed by obscure, rounded, concentric ridges, which give to the former a subnodulose character; intervals marked with small circular pits, probably the points of attachment for spines. Dorsal valve unknown.

Dimensions.—Length of ventral valve, 1.40; width, 1.48; height, about 0.59.

This shell is very interesting as no species of the genus has heretofore been observed in American strata. In Europe it has not been found below the Permian.

Geologic position and locality.—White limestone of the Guadalupe Mountains, New Mexico and Texas.a

---

aTrans. Acad. Sci. St. Louis, vol. 1, 1856-1860, p. 292.

---

There can be little doubt that the fossils in hand belong to Shumard's species, which seems to be fairly common in the Guadalupian fauna, since four examples, none of them, however, very perfect, have come to hand. Shumard's description, together with the figures, will afford an adequate notion of this form and I find it unnecessary to make many additions or changes of moment. The obscure, rounded, concentric ridges, however, mentioned by this author are certainly not obvious on my specimens, the ornamentation of which consists of short, slender elevations like interrupted ribs, which are often more prominent at their anterior end and sometimes carry short (?) spines. The arrangement of the ridges into longitudinal or diagonal rows is not very striking or persistent. The intervals between frequently contain little pits or dimples from which spinules project into the inside of the shell, and these are represented on internal molds as small holes somewhat irregularly distributed. The surface ornamentation is not unlike that ascribed to Productus norwoodi, especially in imperfect conditions of preservation, and as Shumard cites that species from the "dark limestone" of the Guadalupian it is possible that the specimens so identified may have really belonged here.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

AULOSTEGES MEDLICOTTIANUS var. AMERICANUS n. var.

Pl. XII, figs. 16 to 16b.

Shell small, transverse, semicircular. Hinge line slightly shorter than the width in front. Convexity regular. Area moderately high. Delthyrium apparently very wide. Beak small and undefined. Dorsal valve not known.

Shell substance thick, marked by fine liræ, about eight in 2 mm., which become indistinct on the sides near the hinge line, and by concentric irregularities of growth, which are especially abundant and strong near the margin. The shell substance, even on the area, is pierced by numerous small holes, more or less regularly distributed, from which probably proceeded hollow spines.

This form has so distinct an aspect and is represented by so fair a specimen that after some hesitation I have given a description and figures of it. Yet its affinities are so uncertain that I would on this account have refrained from doing so. The fact that the apex is partially broken away might indicate that this was not an Aulosteges but an attached form like Strophalosia, but the other characters are less like Strophalosia than Aulosteges. The uniform size and distribution of the perforations would seem to indicate that they were an original feature, yet no trace of spines has been observed. They may have been produced by some boring animal, a hypothesis for which the surprising thickness of the shell might be pointed to as evidence. In that case the form would probably be a Streptorhynchus, as no internal plates are present.

This species possesses certain points of marked similarity with Aulosteges medlicottianus Waagen, and yet there are some important differences, since it is smaller, without a sinus, longer at the hinge line, apparently with a wider delthyrium, and without ears.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

AULOSTEGES MAGNICOSTATUS n. sp.

Pl. XXXI, figs. 4 to 4b.

The typical specimen is diminutive in size, but from the strong convexity probably represents a mature shell. The length of the ventral valve is 9 mm., the length of the dorsal valve 7 mm., and the width at the hinge line, which exceeds that below, 8 mm.

The ventral valve has an extended instead of an incurved beak, which gives it a shield-shaped outline. The posterior portion consequently is flattened longitudinally, rising gradually toward the front, where, and at the sides, it is strongly deflected. The area in the typical specimen is about 2 mm. in height, flat, and symmetrical. The pseudodeltidium is narrow.

The dorsal valve is transversely semielliptical in shape, flattened over the visceral area, strongly curved at the front and sides. Beak small, inconspicuous.

The surface of the dorsal valve is marked over the visceral area only by fine, somewhat irregular, concentric wrinkles. Toward the front and sides the wrinkles die out and are replaced by relatively coarse radiating costæ, which are faint at first but become strong later.

The ventral valve is more or less broken and the sculpture lost. It probably agreed with the dorsal valve in this particular, as it is known to do over the more peripheral portions, where it is marked by coarse, strong costæ. In the typical specimen both valves, but especially the ventral, seem to terminate in a narrow, concentric, raised band, in whose elevation the costæ lose themselves, so that the band is smooth, except for a very distinct row of coarse spines, one for each costa. There is a row of such spines along the margin of both the dorsal and ventral valve.

The singular feature last mentioned is certainly a character of the specimen under description, but it is as yet doubtful whether it can be regarded as a specific character or only an individual peculiarity.

In its general expression this form resembles certain shells which have more or less of the appearance of Productus semirecticulatus, but have been placed by different authors in the genus Strophalosia. Productus leplayi is one of these, but in this case at least the generic position, I judge, is almost certainly with Productus, and it may possibly be the same in the others.

Horizon and locality.—Delaware Mountain formation, mountains northwest of Marathon, Tex. (station 3840).

AULOSTEGES sp. a.

This species is based on two specimens whose original position was in the black limestone at the base of the Guadalupe section. They appear to be closely related to Aulosteges medlicottianus var. americanus, but their widely different stratigraphic position and faunal association are presumptive evidence that they belong to a different species, and, in fact, certain specific differences are indicated, though a satisfactory discrimination is not at present entirely possible, owing to their imperfect condition. They are smaller than the figured specimen of Aulosteges medlicottianus var. guadalupensis, but their general shape is nearly the same. They have a thick fibrous shell perforated by many small rounded tubes and apparently marked externally by rather fine radiating liræ. It thus appears that they are very similar to the typical specimen from the Capitan limestone, but the perforations or tubes are much smaller and more numerous. Owing to the imperfect condition of these specimens, however, it is possible that they are only morbid examples of Enteletes sp. c, which is found associated with them.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2967).

AULOSTEGES sp. b.

This form is represented by a single specimen from which the shell has been largely removed. It is a ventral valve. The shape is subcircular. The greatest width is 18 mm. and the length of the aperture 14 mm. The hinge line is a little shorter than the width in front. The area is flat, sharply defined, inclined backward at rather a strong angle, and 6 mm. in height. The pseudodeltidium is narrow (1.5 mm.) and strongly elevated.

The surface is marked by rather faint concentric undulations, which can be seen on the mold of the interior and where the shell is retained at one side near the hinge, by somewhat irregular, fine, lamellose concentric lines. The area is profusely covered with perforations left by the hollow spines. So far as can now be told, these are much more rare on the remainder of the shell.

From the facts at present obtainable, this form is of a somewhat different type from those recognized in this report, but I hardly feel justified in view of the poor condition of my specimen and the imperfect data which it furnishes in giving it a distinctive name.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe. Mountains, Texas (station 2967).

Family RICHTHOFENIIDÆ Waagen.

Genus RICHTHOFENIA Kayser.

The genus Richthofenia has been found in only a few limited areas, but these taken together constitute for the genus a wide distribution. We have one species in the Salt Range of India and another at Lo Ping, in China. Gemmellaro's papers are, unfortunately, in part inaccessible to me, but I learn from Schellwien that he has identified Richthofenia among the fossils from Palermo in Sicily. Lastly, the appearance of a species belonging to this genus in southwestern United States lends to the Guadalupian fauna a feature of peculiar interest.

While in most essential respects and in many trivial ones Richthofenia permiana is in close agreement with the Indian and Chinese species, one or two differences appear to be of considerable moment. In R. permiana there is a much less extensive development of the cystose tissue, but this may probably be reckoned a character of specific rather than generic difference. The feature which I had especially in mind in the foregoing comment, however, was the absence of the three structures which Waagen calls septa, extending along the inner surface of the area in the ventral valve. R. permiana shows what may be called two dental callosities or ridges, which occupy the position of the two lateral septa of Waagen's shell, but they are by no means of such a character as would warrant calling them septa. This term is much more appropriate to the structures represented in Waagen's figures. The median septum of Waagen's shell, on the other hand, seems to have no corresponding part in the Guadalupian species. There is, it is true, a sort of longitudinal ridge or septum on the opposite or anterior side of the ventral valve, and it is possible that thin sections through the lower portion would show that this ridge was connected with the development of a median septum, as in the Indian species, and that it might also bring to light two corresponding lateral septa possibly embedded in the cystose tissue; but the available material was too scanty and imperfect to warrant consuming it for sections. In natural sections and exposures, however, these three septal structures appear to be entirely absent from R. permiana.

The median septum seems to play a much more important part in the structure of the Sicilian species of Richthofenia, and still more in Scacchinella, which Schellwien regards as a related genus. Scacchinella occurs with Richthofenia in the Sicilian fauna, and appears to replace it in that of the Carnic Alps, where Richthofenia, so far as known, is missing. A species from Malla Sangcha, in the Himalaya, has been referred to Scacchinella by Diener, but from his figures and description I can not but feel that the reference is a very doubtful one. The genus Scacchinella therefore, so far as at present known, is restricted to the three or possibly two localities above mentioned, while Richthofenia forms a rather striking bond between the Guadalupian fauna and those of the Salt Range, of Lo Ping, and of Palermo. Of course there is nothing at all comparable in the Carboniferous faunas of the Mississippi Valley.

Waagen regarded this genus as representing a distinct family, the Richthofeniidæ, and a distinct suborder, the Coralliopsida; but in this I believe he has somewhat overestimated the really remarkable characters of this type. Most subsequent authors, I believe, have failed to recognize Waagen's suborder. Schuchert places the Richthofeniidæ beside the Productidæ, which is in accordance with my own view. Hall and Clarke include Richthofenia among the genera incertæ sedis. Schellwien seems in doubt whether the greatest affinities are with the Productidæ or with the Strophomenidæ. In weighing the evidence he omits one factor, which, though present in the American and Indian shells, may be absent from the Sicilian species—the development of tubular spines as a part of the surface structures. This is eminently a feature of the Productidæ, in distinction from the Strophomenidæ. Indeed, speaking now solely of the American species, I seem to see in Richthofenia permiana a distinct relationship with Productus, through Strophalosia and Aulosteges. The general character of the sculpture—spiniferous, without ribs, but with strong growth lines—immediately recalls certain Producti; the high area and pseudodeltidium are found in Aulosteges, while the little ridge which stands opposite to the area on the inside of Richthofenia finds, in some cases at least, an apparently analogous structure in Strophalosia. (See the figures of S. cornelliana on Pl. IV.)

In the Sicilian form, as described by Schellwien, both in its own structure and in its relation to Scacchinella, one might be pardoned for believing he found a different generic type, perhaps much more closely allied to the Strophomenidæ. It appears, in the first place, that instead of having two dental callosities, as in the American form, and to some extent as in that from India, the Sicilian type has a single large pyramidal column, the axis of which is a good-sized median septum. This solid column extends along the back of the valve and well down toward the apex, and contains embedded in it a sort of primordial shell, the like of which is entirely unknown as yet in other species. It would appear also, from the fact that they are not mentioned in Schellwien's description, that the surface was not furnished with spinous projections, but I have observed what I regard as the scars of small spines on the surface of species of Scacchinella from the Trogkofelschichten. Thus the Sicilian shell departs rather widely not only from the Guadalupian but from the Salt Range species. The latter has a small septum, according to Waagen, the greater development of which in the Sicilian form may have given rise to the modification of the cardinal structures observed in it. It occupies in some respects an intermediate position between the Guadalupian and the Sicilian species, having the septum of one and the spinous surface of the other; but on the whole it seems to me that Richthofenia permiana is closer to R. lawrenciana, and therefore to typical Richthofenia (though R. sinensis should really be considered the type species), than is the Sicilian form.

The Richthofeniidæ evidently afford an attractive field for further research in case other species or better preserved material are obtained, especially if the different forms supposed to belong to the group could be collated.

The little specimen represented by figs. 10 and 10a of Pl. XXIV seems to be a young example of this genus and to throw some light on the early stages of development. The growth, instead of being regularly conical, appears to be schematically that of a strongly curved triangular plane bent around so that the two edges meet in a sort of suture, indicated more by deflections of the growth lines and by a notch in the upper margin than by a callosity. Apparently at a very immature stage, it may be almost immediately after cementation, by a luxuriant development of the mantle of the ventral valve, the shell was so deposited that, the sides extended backward until they met behind, thus enveloping the posterior portion of the valve of which they form a part, which instead of being without an area, as in Productus, was prolonged upward with the lofty area and normal pseudodeltidium of Aulosteges. The intervening portion between the area and the enveloping sides was filled in with cystose shelly matter, which in certain conditions of preservation peels off, leaving an area and pseudodeltidium much resembling the productoid genus above mentioned.

It is possible that the little shell which suggested this explanation of the origin of some of the singular features of Richthofenia does not really belong to that genus, although I believe that it does; but such a contingency does not necessarily invalidate the explanation itself, which is a little more specific than any I have seen offered elsewhere, though not differing from them in principle.

If the specimen in question does not belong to Richthofenia it would probably go with Tegulifera, a genus which Schellwien introduced as a member of the Productidæ. Tegulifera appears to have been developed by much the same process which I have hypothetized for Richthofenia, except that the area remained narrow in Tegulifera, which accordingly stands in about the same relation to Productus that Richthofenia as here interpreted does to Aulosteges.

Schellwien places Tegulifera in the Productidæ, but apparently does not believe that Scacchinella and Richthofenia have such relationship. His doubts in regard to these forms from the Trogkofelschichten may be justified, and while I seem to see on the part of my Guadalupian representatives of Richthofenia a greater resemblance to the Productidæ than is found in those from the Trogkofelschichten, it seems inadvisable to place Richthofenia and Tegulifera in the same family.

RICHTHOFENIA PERMIANA Shumard.

Pl. XIV, figs. 27 to 27d; Pl. XX, fig. 23; Pl. XXII, figs. 6 to 6b; Pl. XXIV, figs. 10 and 10a; Pl. XXXI, figs. 1 to 3.

1859. Crania Permiana. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 395 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

Shell large, somewhat variable in form, upper or larger valve very much elongated, subconical, expanding most rapidly toward the base; transverse section elliptico-subquadrate; anterior side flattened; sides rounded and in some specimens subangulated posteriorly; posterior side flat or very gently convex, showing, when the exterior crust is exfoliated, a long, narrow pseudodeltidium marked with distinct arched striæ; apex nearly or quite marginal, obliquely truncated and sometimes excavated in front. Surface marked with concentric lines, crossed by irregular, interrupted longitudinal (not very distinct) rugæ, which are most prominent, and assume a somewhat varicose appearance, near the base. Smaller valve very gently concave, surface markings obliterated. Interior characters unknown.

Dimensions.—Height of larger valve about 1-1/2 inches; width at base, about 1 inch.

Locality.—White limestone, Guadalupe Mountains.

I have very little doubt either that the material before me belongs to the species for which Shumard proposed the name Crania Permiana, or that it actually belongs to Kayser's genus Richthofenia. Examples belonging to this genus are represented from a number of localities and horizons, and though all have been referred to a single species, since I am uncertain in what characters and within what degrees to fix the limits of specific variation, it will be better to give a brief notice of each occurrence, both as to superficial and as to structural characters.

Shumard describes R. permiana from specimens obtained in the white limestone of the Capitan formation, and our fossils from that horizon can therefore be regarded as representing the typical form of the species. The shape is generally conical, but much varied as to detail. The transverse section ranges from circular to oval, and one side or another may be more or less flattened. These specimens are not greatly distorted, but their rate of expansion varies in different individuals, and often in the same individual at different places. Thus often arise transverse ridges, which are very irregular and seldom extend through a complete circumference. Scattered over the surface are here and there perforations which originally extended as hollow stolonous spines. The largest specimen measures 25 mm. in length, and is incomplete at its base. This is the condition of many specimens, and doubtless resulted from their habit of attachment. The broken apex seldom fails to show the characteristic cystose structure. In the Indian species the conical interior is subdivided by a transverse cystose partition. In these American specimens the cysts appear to occupy the apex of the cone. If there was a chamber below them it must have been small.

The opercular upper valve is represented by several examples. It is flat and marked by very delicate concentric striæ. One specimen especially shows its position and character. The ventral valve in this case is elliptical in transverse outline, and distinctly flattened on the cardinal side. The dorsal valve is elliptical and smaller than the external circumference of the ventral, the inner wall of the ventral shell being separated from the outer by the cystose layer. This layer is rather thin along the front, but increases irregularly toward the cardinal line, where the greatest thickness occurs. Here, however, it is cut rather squarely off, to form a broad longitudinal groove, into which fits a quadrate projection from the opercular valve. This projection, which does not expand at the cardinal line, as in the Indian species, reaches almost to the outer wall, or, rather, the inner wall of the ventral valve retreats at this point almost to a union with the outer one.

From the "dark limestone" seven specimens have been obtained. They show excellently the superficial characters, the unequal growth, the concentric striæ, and the tubular spines. Two examples are partly exfoliated toward the apex and exhibit the structure which Shumard with much acumen calls the pseudodeltidium, in specimens similarly preserved. What, considered from the inside, appears as a deep, broad groove, in the exfoliated condition is a corresponding ridge which bears along its center another much narrower but strongly elevated ridge, having the appearance of a pseudodeltidium. No groove corresponding to this was observed in the other specimen described in some detail, but this structure seems to be restricted to the portion of the ventral valve below the opercular valve.

Practically the same features are shown by a specimen from the Delaware Mountain formation—the only one found—which occurs as an internal mold. The posterior or cardinal side is flattened and bears a central broad, flattened ridge, which dies out toward the apex. This ridge in turn has a strongly elevated, narrow central ridge, which disappears also toward the apex, toward which it does not extend as far as that upon which it is borne. It is probable that both would likewise die out before reaching the mouth of the shell if the specimen were complete.

In the black limestone at the base of the Guadalupe section three specimens were obtained, one mature and the others very small. No internal structures are shown, but externally they have all the characters of Shumard's species. The two small examples, one of them especially, show a rather significant feature, namely, a distinct notch on what is probably the posterior side of the upper margin, accompanied by a corresponding deflection of the growth lines. This apparently indicates the union of the lateral portions of the shell which have been prolonged backward so that they meet behind the area, thus enveloping this posterior portion. It has not been possible to make observations on all my specimens in this particular, but this deflection of the growth lines is certainly found in other specimens, though some appear to be without it. That the characteristic internal structures are developed at an early period is shown by some very young specimens in our collections. Cystose tissue seems at this stage to be entirely lacking, but the dental callosities have already appeared. They are small and delicate, however. The occurrence of this significant form at this lowest horizon known in the Guadalupe Mountains is important as indicating the thickness of the Guadalupian series at this point.

A few specimens have also been obtained from the Glass Mountains, and as they are silicified a favorable opportunity is afforded for studying their internal structure. One somewhat imperfect example must have had a length of 50 mm. The chamber of habitation extends almost to the base in these examples, as it appeared to do in those from the Guadalupe Mountains, the cellular tissue being confined to a small amount on the back and sides. The cardinal side is especially thickened in this way. Here, as has been described from other specimens, is excavated a rather deep, broad groove having another narrow groove along its median line. On either side the shell stands out in a ridgelike projection. At a certain point 7 mm. below the top in the large specimen previously mentioned, the cellular tissue abruptly ceases, producing a sort of circumferential shelf or groove, probably the position of the smaller valve. Opposite the cardinal groove a longitudinal thickening of the shell occurs, varying in extent in different specimens but forming in every case an appreciable ridge. The inside of the shell below the opercular valve is uneven, though fairly smooth, the chief feature of mark being the presence of a few small tubes parallel to the wall and partly sunk in it, the upper ends of which are open and directed toward the aperture. These without much doubt are connected with the hollow tubular spines. Above the position of the dorsal valve the interior of the shell is rough, being pustulose and pitted. The pits are the same size as the perforations of the spines, but apparently they do not extend to the outer surface. The general character of the outer surface has already been described. Only this remains to be added, that above the plane of the dorsal valve the perforations caused by the tubular spines appear to cease, only irregular lines and ridges being present. The three vertical septa described by Waagen have not been observed, and it hardly seems that they could have been developed. The pallial impression has apparently not been retained on my specimens, though Waagen noted it in his Indian shells.

From the southern Delawares (station 2969) about ten specimens and fragments have been obtained. They agree in most respects with the material gathered elsewhere. The chief difference is that these shells tend to be a little more slender and to develop a somewhat greater amount of the cystose tissue, which now deeply fills the lower portion and, extending up around the sides in a thicker layer, chokes the cavity into a more contracted chamber. I do not say that this is true in every case, however.

I have found it impracticable to distinguish more than one species in the material examined without assigning considerable importance to characters which seemed either more or less trivial or else very variable. On the other hand, with the characters said to be present in the Asiatic species but apparently lacking in this, and those present in this but apparently lacking in the other, there can be but little doubt that R. permiana is distinct from R. lawrenciana and R. sinensis. In any event, the American form antedates either of the names just mentioned. In fact some of the structural differences which appear to subsist between the American and Asiatic forms have the appearance of being so important as to suggest the consideration whether they really belong to the same genus. The general appearance is so strikingly similar, however, and the type of structure so peculiar in every way, that I would prefer to reserve the consideration of this point until a wider knowledge of the specific and generic variation will afford a juster conception of the importance which must attach to different lines and degrees of variation. There seem in fact to be but scant possibilities for variation in these forms except in rather trivial or else rather essential characters, and it may be that differences of structure such as exist between the Indian, the Alpine, and the American representatives of the group will have to be used for specific rather than generic discrimination, as would perhaps be the case in other types.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930) and Guadalupe Point (stations 3762b and 3762e); Delaware Mountain formation, Guadalupe Point (station 2931); basal black limestone, Guadalupe Point (stations 2920 and 2967), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2964 and 2969). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Family ORTHIDÆ Woodward.

The Orthidæ are a declining group in the later horizons of the Carboniferous, showing but limited variation along generic lines and represented often scantily in individuals. Enteletes, the most characteristic and abundant representative of the family at these horizons, is the only type which has been found in the Guadalupian fauna, though other genera occur in some of the faunas with which comparison has been made. In the Guadalupian, indeed, this genus is restricted, so far as known, to the earlier portion of the series, none of the Orthidæ having thus far been found in the Capitan formation.

Seven species of Enteletes are recognized in the present report, all of them, with one exception, probably belonging to the commoner division of the genus, the ventrisinuati.

In India, in the Salt Range, Waagen recognizes seven species of Orthis, which may now be distributed into the groups of Rhipidomella, Schizophoria, and Orthotichia, none of which, of course, is represented in the Guadalupian fauna so far as known. Of Enteletes he finds two species of the ventrisinuati and five of the dorsisinuati. E. kayseri. is perhaps comparable to E. dumblei, at least in configuration, while E. lævissimus may be compared with Enteletes sp. e. The dorsisinuati, on the other hand, contain some singular and striking types, such as E. pentameroides and E. latisinuatus, but nothing which closely resembles the Guadalupian E. globosus.a The Indian fauna thus comprises a number of orthoid genera not found in the Guadalupian and a rich diversity of the dorsisinuate type of Entetetes, while it contains but two rare species of the ventrisinuati, a condition very nearly the reverse in every way of that which, at the time of writing, appears to exist in the Guadalupian fauna.

---

aIt may be noted that practically all the types with extreme gibbosity, large plications, and generally extravagant development seem to belong to the dorsisinuates. This group, as before remarked, is almost absent from American faunas, and this, the only known species, is relatively modest and normal in its development of characters.

---

In the Himalayan region, at Chitichun, Diener found a species of Enteletes, one of the ventrisinuati, which he described as new, under the title Enteletes tschernysche. It is of the same general type as E. dumblei and Enteletes sp. d of the Guadalupian. In his second paper on this fauna Diener cites E. tschernyschewi, E. waageni Gemm., and E. subaequivalvis Gemm., E. waageni comprising part of the original material of E. tschernyschewi. E. tschernyschewi and Enteletes cf. elegans Gemm. are cited from Malla Sangcha. All of these forms appear to be closely allied to the two Guadalupian species mentioned above. Salter also cites Orthis sp. from Niti Pass, and Davidson Orthis sp. from Tibet.

From the Carboniferous of Turkestan Romanowsky figures a small Orthis, probably incorrectly identified as O. resupinata, but cites no Enteletes and nothing similar to the Guadalupian orthoids.

In the fauna from Lo Ping, in China, we have a species of Orthis and one of Enteletes, which Kayser identified with Rhipidomella pecosi and Enteletes hemiplicatus, respectively. The Chinese forms occur associated with Richthofenia, Leptodus, and a generally different fauna from that of the Pennsylvanian of the Mississippi Valley, where in America these species are most commonly found. E. hemiplicatus Kayser, however, is considerably different from the American E. hemiplicatus, and Waagen, recognizing this fact, gave to the Chinese species, which he identified also in India, the name E. kayseri. The identity with Marcou's species of the form from Lo Ping called by Kayser Orthis pecosi had been questioned by both Waagen a and Schellwien,b and in 1901 Fliegel introduced for it the specific name subquadrata. E. kayseri is a member of the ventrisinuate group, and the typical form from China more closely resembles E. hemiplicatus than does the Indian E. kayseri figured by Waagen. Of the Guadalupian species, E. dumblei is comparable to the Chinese form in general configuration, though clearly a distinct species. Enteletes sp. c is probably still more closely similar, though it is doubtful whether when of the same size as Kayser"s figures it would have had equally strong plications.

---

aMem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 574.

bPaleontographica, vol. 39, p. 35.

---

Among the numerous more or less imperfectly known faunas described from China by Loczy, orthoids are for the most part absent. The only instance where the group is represented is in the collection from the vicinity of Kantschoufu, where Loczy cites an unnamed species of Orthis which he compares to O. lyelliana (non De Koninck) of the Moskovian of Russia and to O. pecosi of the Pennsylvanian of the United States, and an Enteletes which he identifies as E. lamarcki of the Russians Moskovian. The latter is more nearly related to a variety of our Pennsylvanian E. hemiplicatus than to any species yet known from the Guadalupian.

The report by Schellwien on the paleontological collections obtained by Professor Futterer in an exploring expedition through Asia has not come to hand, and the only account of these faunas known to me is a short paper by Schellwien, which contains no citations suggestive of the Guadlupian fauna.

In the fauna from Padang, in Sumatra, described by Fliegel, occur two orthoid types, which he describes, one as new, under the title of Dalmanella frechi, the other provisionally assigned to Dalmanella michelini. The former species, which alone is figured, is probably an Orthotichia, an opinion which is supported not only by its configuration but by what is said of its internal structure (reported to possess in both valves a median and two lateral diverging septa), and by the fact that it is stated to be closely related to Orthis derbyi Waagen. Nothing similar to these forms, of course, is known in the Guadalupian.

From the Permian beds of Timor and Rotti, in the Indian Archipelago, no orthoids have been cited, so that a comparison with the Guadalupian fauna is impossible in point of this branch of the Brachiopoda, but the absence of such types is a matter of agreement with the fauna of the Capitan formation of the Guadalupian.

Roemer figures a large orthpid from the west coast of Sumatra, which he identifies as Orthis resupinata. It would appear to be a Schizophoria or an Orthotichia, and has no allied Guadalupian form.

Among the Orthidæ of the "Permo-Carboniferous" of Queensland and New Guinea Etheridge records only two species of Schizophoria, a type which is unknown in the Guadalupian. It is surprising to find Enteletes absent from the Australian faunas.

The Orthidæ discussed in De Koninck's account of the Carboniferous faunas of New South Wales comprise only two species—O. resupinata and O. michelini. Both species seem to belong to the lower horizon and are not germane to the present consideration.

The so-called Lower Carboniferous of the Russian section, comparison of which with the Guadalupian can be hoped to yield but little profit, being disregarded, the next succeeding formation is the "Middle Carboniferous" or Moskovian. It is in this division that I seem to see especially close faunal relations with the Pennsylvanian of the Mississippi Valley, perhaps more with the lower portion than with the upper. This is the horizon of occurrence of Enteletes lamarcki, which appears to be chiefly allied to our own E. hemiplicatus. E. lamarcki may be described as a ventrisinuate Enteletes which has an incipient plication on the fold and sinus, and may thus possibly be regarded as a passage form to the dorsisinuati. Shells having this configuration are also accompanied by others which have a similar shape but lack the mesial plication, and are thus true ventrisinuates. The same condition occurs in E. hemiplicatus, but there the true ventrisinuate type is dominant, while in Russia that having a mesial plication appears to be most characteristic. No Guadalupian species yet known possesses the latter character, but with the other variety, E. dumblei, and, to a less degree, E. angulatus, show some general similarity. In addition to Enteletes, the Moskovian fauna contains species of Rhipidomella, and also of Schizophoria.

From the Gschelian stage, which next succeeds, Nikitin cites Enteletes lamarcki and another species of Enteletes which is not identified. In Tschernyschew's recent work on the "Upper Carboniferous" brachiopods of the Urals and Timan, in which the term "Upper Carboniferous" appears to be used as exactly equivalent to "Gschelian," he cites no species of Enteletes, a condition of affairs which is similar to that prevailing in at least the Capitan formation of the Guadalupian; but he recognizes two species of Rhipidomella, two of Schizophoria, and one of Orthotichia, types which are thus far not known in the Guadalupian and may indicate an earlier age for the Russian beds.

The fauna of the Artinsk ("Permo-Carboniferous") and that of the true Permian contain, as is well known, but relatively few brachiopods, and among them no representatives of Enteletes. From the Artinsk, however, Tschernyschew figures a species of Schizophoria, which he designates as Orthis cf. O. indica, and Stuckenberg, while recording no representatives of the family from the Artinsk, notes Schizophoria resupinata in the accompanying Kungur beds. Krotow also cites Schizophoria resupinata from the Artinsk, but from the Russian Permian the family is, so far as I am aware, unrepresented.

From Armenia Abich cites of this group only a small form, which with doubtful propriety he identifies as Orthis resupinata. This fauna, therefore, affords nothing which invites comparison with the Guadalupian orthoids. The same is true of Rhipidomella michelini, which Frech identifies from the upper "Lower Carboniferous," and Schizophoria indica, which Arthaber cites from the "Upper Carboniferous" of the same principality.

The fauna of Balia Maaden, in Asia Minor, contains but one orthoid, according to Enderle, who identifies it as Orthis aff. resupinatæ Martin. Whether this form is a Schizophoria, as indicated by Enderle, or an Orthotichia, does not concern us. It is remote from the Guadalupian orthoids at present known.

In his work on the Fusulina limestone of Palermo, Gemmellaro mentions no species of Schizophoria, Orthotichia, etc., but describes no less than ten of Enteletes, all belonging to the ventrisinuate group, except possibly E. microplocus, the small size and considerable number of whose plications do not permit it to be assigned definitely. In this particular, accordingly, the Sicilian fauna is comparable to that of the Guadalupe Mountains, and a number of species are analogous.

In the character of its species of Enteletes, as well as in its apparent absence of other types of Orthidæ, Gemmellaro's fauna is especially like the Guadalupian.

In the Carnic Alps Schellwien discriminates one species of Schizophoria and seven of Enteletes,a of which four belong to the ventrisinuati and three to the dorsi sinuati. The latter group is much better represented than with us, and contains the striking species Enteletes suessi, but nothing very similar to our E. globosus. The ventrisinuates do not compare very closely with Guadalupian species of the same group.

---

aAs noted below, Schellwien includes Orthoichia with his Enteletes.

---

The Dyas of Germany and the closely related Permian of England appear to be without representatives of this family, and agree in this particular with the typical Russian Permian. Although the Guadalupian contains but little which positively suggests these faunas, the Capitan formation is in agreement with them in lacking this group of brachiopods.

From the Carboniferous of Spitzbergen nothing is yet known which can be compared with the orthoids of the Guadalupian fauna. It is true that from the south point of Spitzbergen Toula cites O. keyserlingiana and from Axel Island Orthis resupinata, of which the latter, from his figures, is suggestive rather of a Spirfer or an Athyris; but these types, so far as known, are not encountered in the Guadalupian. Orthis eximiiformis, described from Nova Zembla by the same author, as indicated by the nomenclature which he himself uses, would, in our present classifications, be a Meekella; but a restoration of his imperfect specimen somewhat different from that which he adopts might make it an Enteletes.

Stache records three small species of Orthis from the West Sahara (Igidi), but they resemble nothing in the Guadalupian, and the associated fauna is probably much earlier. De Koninck cites Schizophoria resupinata and Rhipidomella michelini from New South Wales, but the genus Enteletes seems not to occur there.

From Bolivia D'Orbigny described two species of Enteletes (E. andii and E. gaudryi), which from his figures are only in a general way comparable to the ventrisinuate species of the Guadalupian. He also describes Orthis cora (probably a Rhipidomella) and Orthis buchii, neither species having, so far as known, a corresponding one in the Guadalupian. Toula described Orthis resupinata var. latirostrata from the same country, the correct generic reference probably now being to Schizophoria. Salter cites Orthis resupinata? and Orthis andii also from Bolivia. His figure of the latter species resembles our common Pennsylvanian Enteletes hemiplicatus and several of the Guadalupian species much more than the original figures of D'Orbigny. Gabb cites Enteletes andii from Peru.

Enteletes was not found by Derby in Brazil, but he cites two unplicated orthoids as O. penniana and O. morganiana, the latter being now the typical species of Hall and Clarke's genus Orthotihcia. Neither of these has, so far as known, any related Guadalupian species.

In the typical Pennsylvanian the Orthidæ are represented by three genera, each containing a single species: Rhipidomella pecosi, Schirophoria resupinoides, and Enteletes hemiplicatus. The first two types are, so far as known, unrepresented in the Guadalupian. Between the Enteletes of the Guadalupian and Pennsylvanian there are some important if not striking differences. At least one of the Guadalupian species belongs to the dorsisinuate section, which has no Pennsylvanian representatives. Among the ventrisinuates also certain points deserve comment, such as the greater differentiation of the Guadalupian types and the fact that none of them can strictly be identified with the Pennsylvanian E. hemiplicatus.

Genus ENTELETES Fischer de Waldheim.

Waagen has made a division of the Enteletes into two groups,a one of which he calls the dorsisinuati and the other the ventrisinuati. In one of these groups, as the name indicates, the dorsal valve bears a sinus, and in the other the ventral. To the ventrisinuati belongs our common E. hemiplicatus Hall. Of the dorsisinuati no North American species have up to the present been found. To this statement some exception must be made, in that Waagena refers to the dorsisinuati a shell from Nebraska City figured by Geinitz as Rhynchonella angulata; but I feel satisfied that the form figured by him is our common E. hemiplicatus, and either that some misunderstanding has produced the present confusion, or that Geinitz uses the terms "kleinere Schale" and "grossere Schale" as equivalent to dorsal and ventral, irrespective of the fact that in Enteletes the dorsal valve is usually the larger.

---

aSalt Range fossils: Mem. Geol. Survey India, Pal. Indica. ser. 13, vol. 1, 1887, p. 552.

---

Schellwien b points out another and more probable instance of a dorsisinuate form in the Mississippi Valley Pennsylvanian faunas in a shell which has several times been figured as E. hemiplicatus, but differs from the typical phase of that species in having an incipient plication on the fold and sinus. The fact of the occurrence of this partially developed plication in no wise obscures the real character as such of the fold and sinus, so that from this point of view the form in question would still remain a ventrisinuate type. While, on the one hand, it is possible to conceive of the dorsisinuates and the ventrisinuates as having sprung directly from a common orthoid prototype (Orthotichia?), which developed a primary fold in one instance on the dorsal valve and in the other on the ventral, the dorsisinuates may have come indirectly through the ventrisinuates by the development of a median plication on the fold and sinus. This has already been suggested by Waagen.c If this inflection of the fold of the ventrisinuate shell were carried to such a point that it equaled in size or exceeded the lateral plications, the real character of fold and sinus would be obscured and the ventrisinuate shell would become dorsisinuate, not to be distinguished from dorsisinuates derived directly from the unplicated shell, if there are any such. Many circumstances tend, however, to confirm the opinion that the dorsisinuates are derived from those having a ventral sinus, in the manner above indicated. In the first place, Schizophoria and Orthotichia, from which Enteletes is supposed to have been derived, comprise shells which with the reversal in dimensions of the two valves have, so far as I have been able to ascertain, the sinus and fold, where these structures are appreciable, on the ventral and dorsal valves, respectively. There is no orthoid, so far as I am aware, which by its characters, external and internal, might present claims to be the ancestor of Enteletes, that has a dorsal sinus, and might be the stock from which the dorsisinuate group is directly derived. There is likewise the fact of geologic occurrence; the dorsisinuates apparently coming in only in the late Carboniferous or Permian, while the ventrisinuates appeared earlier. Waagen also calls attention to the significant fact that the geologically oldest species of the Salt Range is the one which is most nearly related to the American "Coal Measures" forms.a

---

bAbhandl. K-k. geol. Reichsanstslt, vol. 16, part 1. 1900, p. 7. footnote.

cSalt Range fossils: Mem. Geol. Survey ludia, Pal. Indica, ser. 13, vol. 1, 1887, p. 562.

aWaagen, W., Salt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, 1887, p. 557.

---

Furthermore, we have in E. lamarcki Fischer de Waldheim and the forms referred to E. hemiplicatus Hall, species representing the very transitionary stage between the two groups. Schellwien has already pointed out that the American form with an incipient median plication can not properly be referred to E. hemiplicatus. It occurs in association with the true E. hemiplicatus, which it so closely resembles in all other characters that I doubt if this new form should be given more than varietal rank.

In regard to the affinities of these shells and their position in the two groups recognized by Waagen, it appears that the latter author refers E. lamarcki to the ventrisinuates,b while Schellwien refers that species, along with the above-mentioned variety of E. hemiplicatus, to the dorsisinuates. Perhaps the logical position would be to refer species to the dorsisinuati on the first introduction of the plication on fold and sinus; but because of the obvious close relationship of one having this configuration to E. hemiplicatus, a ventrisinuate type, and the imperfect development of the median plication, I am disposed to retain them in the ventrisinuati. The fact that there is some divergence of opinion as to the group to which some of the Enteletes types belong indicates how closely related are the two branches discriminated by Waagen, although they appear at first so distinct.

---

bIdem, p. 562.

---

Schellwien includes in Enteletes shells which by other authors are distinguished as a distinct division, under the name of Orthotichia, or in some cases retained as Orthis pars. It will be remembered that Orthotichia was introduced by Hall and Clarke c for the type of orthoid structure characterizing those forms which Waagen d had called the group of Orthis morganiana. The introduction of a new name for this type was in effect the carrying out of a suggestion made by Waagen, that the three groups or divisions of Orthis recognized in his work might be of subgeneric rank. The species after which Waagen named this group or division was by Hall and Clarke taken as the type of their Orthotichia. As pointed out by Waagen, and later by Hall and Clarke, this group is intermediate between Orthis (Schizophoria) and Enteletes; and Schellwien e has merged it with the latter genus for the reason that if Orthotichia be placed with Orthis the only difference distinguishing Orthis from Enteletes is the absence of plications. Schellwien also points out a possible danger in the use of Orthotichia, which, as defined by Hall and Clarke, would include both shells that were in the process of evolution from Schizophoria to Enteletes and also such types of Enteletes as have lost their plicated surface and retain only the characters of the transition forms (Orthotichia proper), though a distinct intermediate stage (Enteletes) exists between them. He also remarks that although Hall and Clarke call Orthotichia a transitional development between Schizophoria and Enteletes, the type species, Orthis morganiana, possesses the characters which he regards as distinguishing the relapsed or atavic type of Enteletes structure.a He suggests that the possession of an inflated dorsal valve and closely arranged septa in the ventral are characteristic of the later forms, the earlier of which he refers to Orthis. He prefers, therefore, an unwieldy or somewhat heterogeneous grouping to one which tends to misrepresent facts of phylogeny. This objection seems to me in some regards to be well taken, for I would be far from advocating a reference to the same genus both of types which were developing into and of those developing out of Enteletes; and if Orthotichia, as based on O. morganiana, does, as suggested by Schellwien, stand for the latter stage, I would favor another name for that which is intermediate in development between Schizophoria and Enteletes.

---

cNat. Hist. New York, Pal., vol. 8, pt. 1, 1892, p. 213.

dOp. cit., p. 564.

eAbbandl. K-k. geol. Reichsanstalt, vol. 16, part 1, 1900, p. 3.

aAbbandl. K-k. geol. Reichsanstalt, vol. 16, part 1, 1905, p. 5.

---

The objection to the use of Orthotichia, which Schellwien advances, seems to be academic, rather than real. According to accepted theory, species which are in process of evolution from Schizophoria (or Orthotichia) to Enteletes would have the posterior portion smooth and the marginal portions plicated. A similar line of reasoning would lead one to expect that in species which are reverting from Enteletes, a condition which Schellwien thinks is manifested in Orthotichia morganiana, the shell would have to pass through an Enteletes stage, so that the posterior or median parts would be plicated and the lateral ones smooth. This condition is certainly not present in Orthotichia morganiana, and I do not recall having seen any specimens or figures illustrating this condition. It is conceivable, however, that instead of passing through an Enteletes stage the reversional tendency might manifest itself in the gradual loss of the power to fold the mantle at all stages of growth. Thus this argument loses much of its force.

Evidence based on the occurrence of species in geologic time presupposes a knowledge of these data and a correlation of beds so much more complete and accurate than that which we at present possess that I am disposed to accept it with much reserve. It may be more or less true according to our present knowledge, as urged by Schellwien,b that the oldest typical Enteletes are strongly plicated, while the faintly plicated forms are especially abundant in Permian strata, yet, at the same time, it is reasonable to believe that smooth and faintly plicated forms must have preceded those with sharply folded shells. In regard to the occurrence of faintly plicated types in the younger deposits of the Carboniferous, I do not see that it is possible to tell whether it is a case of persistence of the faintly plicated stock in association with the strongly plicated one (Enteletes) to which it gave rise, or a persistence of the strongly plicated type with the reverted shells of faint plication which descended from it. Neither hypothesis is at present susceptible of very strong direct evidence, and as the former is more simple I am prepared to employ it until better reasons are adduced for supposing that any of the shells having the appearance of Orthotichia possessed Enteletes as an ancestor.

---

bIdem, p. 4.

---

If I understand Schellwien aright, he proposes to place the intermediate species with Orthis and to merge with Enteletes the relapsed ones, abolishing the name Orthotichia, the type of which, he thinks, belongs to the latter group. My position is somewhat that of an agnostic on the question of there being a relapsed group, and I doubt whether Orthotichia really belongs to it if there is one, but I will go so far with Schellwien as, while recognizing Orthotichia, to place it as a subgenus of Enteletes rather than of Orthis. This comes in practice very near to following Schellwien in this matter without, however, accepting the opinions on which his course of action is based.

In the present paper seven species of Enteletes are distinguished, and they will be referred to as E. globosus, E. dumblei, E. angulatus, Enteletes sp. a, Enteletes sp. b, Enteletes sp. c, and Enteletes sp. d. Of these, E. globosus belongs to the dorsisinuati, and as such is of considerable interest, since Enteletes has up to the present time been known to be represented in North America only by the ventrisinuati type. A minor subdivision of the Guadalupian ventrisinuate species, none of which seem to be identical with our common E. hemiplicatus, can be made, although but provisionally, as so much of the material is fragmentary.

E. angulatus represents one group, and with it should perhaps be associated the well-known form of the Mississippi Valley Pennsylvanian, E. hemiplicatus. E. dumblei would appear to deserve a position in a distinct group by reason of the unusual character of its minute surface sculpture. The little-known types Enteletes sp. a, Enteletes sp. b, and Enteletes sp. d probably belong together, being characterized by having numerous small angular plications. Enteletes sp. c probably represents a group by itself, and is of interest because it apparently belongs to that type of Enteletes which has obsolescent plications and which, appearing in special abundance after the strongly plicated forms had been dominant, represents, as Schellwien thinks, a retrogression of the genus from a plicated to a smooth condition. On the assumption that the Guadalupian specimens are mature, they represent a stage about equal to Enteletes dieneri Schellwien and a little less unplicated than Enteletes derbyi var. demissa Schellwien, However, young specimens of the size of the latter species would have been equally unplicated. If the Guadalupian form really does belong to a retrograde series, its position at the base of the section is surprising, but may be logically connected with the entire absence of the orthoid group in the Capitan formation.

One may not be certain that some, perhaps all, of the shells forming Enteletes sp. c do not belong to the group of Orthotichia, or that, if grown to a larger size, they would not have been distinctly plicated. This uncertainty often arises when the specimens collected are few in number and of small size, and in surveying the literature I have come upon a number of instances in which forms that I was led to suspect might be young Enteletes have been placed with Orthis or Orthotichia. In the present case also the uncertainty as to whether these small shells are young or mature, and, indeed, whether they have as immediate ancestors plicated or unplicated forms, makes it dangerous to speculate on their biologic or stratigraphic significance.

In the Guadalupian section in the immediate vicinity of Guadalupe Peak the orthoid group is surprisingly rare. Fairly abundant in the black limestone at the base of the section, where it is represented by Enteletes sp. c, but one specimen has come to hand from the Delaware Mountain formation overlying (Enteletes sp. d), while in the Capitan formation the genus, and indeed the entire family, is not known to occur at all. In areas contiguous, and in association with faunas in some cases doubtfully and in others satisfactorily correlated with the Delaware Mountain formation, the chief collections by which this group is represented in the present report have been obtained.

There are, in almost every group of fossils, a number of poorly characterized—or one may say primitive—species, which if not actually identical the world over can nevertheless scarcely be discriminated in the different provinces in which they occur. Aside from such forms the Guadalupian Enteletes, so far as known, are more similar to those of the Permian (?) of Palermo than to other groups of species whose descriptions are known to me.

DORSISINUATI.

ENTELETES GLOBOSUS n. sp.

Pl. XXX, figs. 1 and 1a.

Shell rather small, globose. Ventral valve moderately convex. Dorsal valve very convex. Outline subcircular. Surface of dorsal valve marked by a broad sinus of moderate depth, bounded by large plications, on either side of which are three small plications. The plications on the ventral valve correspond, so that there are a median sinus and four lateral plications on the dorsal valve and a median fold and four lateral plications on the ventral valve. The fold and bounding furrows of the ventral valve, and the sinus and bounding ridges of the dorsal valve are larger and more prominent than the other grooves and ridges. The superficial liræ appear to be rather coarse, equal, and regular; but their character has been obscured by silicification. The liræ appear to be considerably coarser than in Enteletes hemiplicatus.

It is only after much hesitation that I have decided to base a new species on a specimen as disfigured as the one described; but it shows many important characters, though some have been obscured. This form, moreover, is one of especial interest since it belongs to the group of Enteletes which Waagen has called the dorsisinuati, and is probably the first representative of this group to be noticed from North America. Fortunately the presence of three septal plates close together in one valve distinguish it immediately as the ventral, while the opposite one bears a median sinus corresponding to the somewhat crushed fold of the septate valve.

This species presents a certain superficial resemblance to Enteletes œhlerti Gemmellaro,a which was first found in Sicily and has subsequently been identified by Schellwienb in the Carnic Alps, but it readily develops from a more careful scrutiny that E. œhlerti belongs to the ventrisinuati, while E. globosus is a dorsisinuate. Among the dorsisinuati I have not found any described species with which E. globosus would be liable to be confused.

---

aGiorn. Soc. di sci. nat. ed econ. di Palermo, vol. 22, 1899, p. 139, pl. 29, figs. 11-15.

bAbhandl. K.-k. geol. Reichsanstalt, vol. 16, part 1, 1900, p. 11, pl. 1, figs. 31-13.

---

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

VENTRISINUATI.

ENTELETES DUMBLEI n. sp.

Pl. XXVI, figs. 4 to 4b.

Shell large, orbicular. Ventral valve moderately convex, subcircular, somewhat contracting at the hinge line; width 26 mm., length about 22 mm. The area is flat, except at the apex, and well defined, making an angle of about 90° with the plane of the shell margin. It is 16 mm. wide, 4-1/2 mm. high, and transected by a delthyrium which is 4-1/2 mm. wide at its base. The beak is small, pointed, and incurved, and it overhangs the area. This valve bears a median sinus, on either side of which are four subangular plications and a fifth smaller and less distinct. The sinus is larger than the sulci between the lateral plications.

Dorsal valve highly inflated. Shape like that of the ventral valve. Beak comparatively large and overhanging the low area. Surface with a median fold and five lateral plications, the outer of which is small and indistinct.

The surface is marked by fine, radiating liræ, just visible to the naked eye. They are thin and ridgelike, though low and faint, the spaces between them being wider than the liræ themselves. About the same number of liræ would occur in a given distance in this species as in E. hemiplicatus; but in the latter they are more distinct, broadly rounded, and with linear interspaces, while in E. dumblei they are thin, narrow, and with wide interspaces.

This shell evidently belongs to the ventrisinuati group of Waagen, which includes also our common E. hemiplicatus, along with a large number of foreign species. Enteletes dumblei can be distinguished from E. hemiplicatus by its broader and higher cardinal area, more numerous plications, and different surface ornamentation. It is related to several foreign species, but to none more nearly than to E. elegans Gemmellaro,b Gemmellaro's figures show a strong resemblance in configuration, but whether this similarity is persistent, and whether the Italian species possesses the rather peculiar striation of the American one, can hardly be determined without a comparison of the shells themselves.

---

bGiorn. soc. di sci. nat. ed econ. di Palermo, vol. 22, 1899, p. 141, pl. 29, figs. 6-10.

---

Horizon and locality.—Hueco a formation, Diablo Mountains, Texas, as reported (station 3764).

---

aInformation obtained after this paper was in type indicates that this species and the following were obtained not from the Guadalupian but from the upper beds of the underlying Hueco formation.

---

ENTELETES ANGULATUS n. sp.

Pl. XXVI, figs. 3 and 3a.

The ventral valve is all that has been obtained of this species. It presents the following characters:

Shell large, transverse, strongly inflated. Width 30 mm., length 25 mm. Beak small, incurved, and overhanging the area. The latter appears to be flat transversely, slightly concave longitudinally. Its width is 17 mm., its height 6 mm., and the delthyrium is 5 mm. wide at its base. There is a large median sinus with three obvious and one obscure plication on each side. The sinus is somewhat larger than the furrows between the plications, and the sinus, furrows, and plications are strong and angular. The more minute surface ornamentation appears to resemble that of Enteletes dumblei, and to consist of narrow, obscure, threadlike liræ separated by comparatively broad interspaces. The convexity of this valve, in view of the fact that it is a ventral, is remarkable. That it is a ventral, however, is clearly shown by the presence of two proximate dental plates and one intermediate septal plate.

This species is especially characterized by its highly arched ventral valve, by the strong, large, angular plications, and also probably by the fine, narrow, and sharp liræ which mark the surface. The character of the liration is not especially noted in many descriptions nor represented in figures, but Enteletes angulatus is distinguished from most of the foreign species which, in the literature, have come to my notice, by the configuration of the shell, as above mentioned. It appears to be most closely related to certain forms from Palermo described by Gemmellaro, especially to E. waageni var. umbonatus.a It must remain for an actual comparison of specimens to determine what are the relations between these two species.

---

aGiorn. Soc. di sci. nat. ed. econ. di Palermo, vol. 22, 1899, pl. 29, figs. 22, 23.

---

Enteletes angulatus should be readily separable from E. dumblei by reason of its fewer, stronger, and more angular plications. I regard it as distinct from E. hemiplicatus, because the ventral valve is more gibbous, the cardinal area larger, the plications stronger and more angular, and extended farther toward the beak. If the surface ornamentation is of the character which it appears to have, this also should furnish an excellent means of distinguishing the two species.

Horizon and locality.—Hueco formation, Diablo Mountains, Texas, as reported (station 3764).

ENTELETES sp. a.

Pl. XXX, figs. 2 and 2a.

This species is represented by a fragment the general character of which is shown by fig. 2 of Pl. XXX. The internal structures of this specimen can not be made out, but the configuration appears to be that of a ventral rather than of a dorsal valve. On this point, however, it is impossible to reach a satisfactory conclusion. The convexity is remarkably slight, if the shell be supposed a dorsal valve. The beak is rather prominent and somewhat incurved, No distinct area can be observed, the curvature being uninterrupted to the point where the shell is terminated by a wide delthyrium. There is a median sinus and on each side of it three distinct plications, with possibly a fourth, which is very faint. The sinus is but little larger than the sulci, and the plications are subangular. The fine superficial liræ are faint, except for rather numerous and equally distributed tubular ones, which are prominent. The others are indistinct, even when viewed with a glass, and are practically invisible without.

If this is a dorsal valve it evidently belongs to the dorsisinuati, and E. globosus. is the only Guadalupian form with which it is necessary to make comparison. From the latter it appears to differ in having the liræ much finer and more indistinct and the convexity very much less. Of the other species of Enteletes here described, in configuration it most nearly resembles E. dumblei, but the different character of the liration distinguishes them at once.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

ENTELETES sp. b.

Associated with the foregoing is a fragment of what probably represents a distinct species. The fragment is nearly square and measures 13 mm. in either direction. There is a median plication somewhat larger than the others. Of the latter, three on each side and part of the fourth are preserved, and the incomplete condition and gradual curvature of the fragment indicate that there were additional ones on portions of the shell now missing. The larger (median?) plication is about 6 mm. wide, the others about 4 mm., and their diminution in size within the limits of the shell preserved is very slight. They are well defined, but neither very high nor very angular. The superficial liræ are fine but strong.

The plications and liræ are those of Enteletes rather than of Meekella; but it is impossible to determine whether the specimen is a dorsal or a ventral valve and consequently whether it belongs to the dorsisinuati or the ventrisinuati. The size of the fragment is considerable and the complete dimensions must have been rather exceptional. The shell representing Enteletes sp. a is likewise so slightly convex that a large complete size is indicated; but the liræ in one case are faint and in the other distinct, so that I have felt that they should not for the present be referred to the same species. The silicification of the form under consideration, however, is coarse and may have exaggerated this character.

This species is probably related to E. microplocus Gemmellaro,a but is certainly distinct from that species by reason of its larger and presumably less numerous plications. The presence of a well-marked mesial plication of larger size than the lateral ones is also a distinguishing feature.

---

aGiorn. Soc. sci. nat. ed econ. di Palermo, vol. 22, 1899, p. 147, pl. 28, figs. 40, 46.

---

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

ENTELETES sp. c.

Pl. XXIV, figs. 4 to 5a; Pl. XXVI, figs. 1 to 2b.

In the black limestone at the base of the Guadalupe section occurs a species of Enteletes which deserves mention, both because of the scanty fauna at present known from that horizon and because of its biologic relations. Though the form is not a rare one no perfect examples have been found and for this reason, and since I am not sure but that certain of the peculiarities exhibited are due to immaturity, I have refrained from proposing a new name for what is probably a distinct and undescribed species.

All the specimens seen are small. The largest size indicated, which several examples approach, is 18 mm. in length and probably but little less than 25 mm. in width. The configuration and striation are practically the same as in E. hemiplicatus. The socket plates are also about as in that species, but there is a low though very distinct median septum in the dorsal valve. The septum and dental lamellæ of the ventral valve do not differ materially from those of the species mentioned. The dorsal valve has a mesial fold and the ventral a corresponding sinus, so that the form can be referred to the ventrisinuate group. A lateral plication on each side (possibly two on the ventral valve) can also be made out, but they are very faint.

The faintness of the plications and the late period in the shell growth at which they appear are the chief features that distinguish this form from E. hemiplicatus in which the plications are much more strongly developed in shells of equal size. The different character of the liration, and to a less degree the obsolescent plications, likewise distinguish it from E. dumblei.

A number of faintly plicated species of Enteletes have been described in the higher Carboniferous terranes of Europe and Asia, in regard to certain of which there may well be some doubt as to whether they are really based on mature specimens. Some of these, like E. sublævis Waagen, belong to the dorsisinuati, but others admit of more or less satisfactory comparison, so far as general appearance is concerned, with the form under consideration. This is especially true of two species from Sicily described by Gemmellaro (E. tschernyschewi and E. Waageni), to one or the other of which Enteletes sp. c may perhaps ultimately be referred as immature or partially developed individuals. This can be ascertained, however, only by comparison of the specimens themselves, to determine both whether the superficial resemblance shown by the figures actually exists, and whether it is borne out by correspondence in the more minute surface sculpture.

A repetition of shells of this size and character in the black limestone would naturally be taken as indicating a mature stage. If so, they might belong either to the group of late Permian species which are supposed to be losing the plicated configuration of Enteletes and returning to a condition similar to Orthotichia, or, on the other hand, to the other group of similar appearance which is in the course of development from the smooth Orthotichia stage to the plicated condition of Enteletes. The horizon of occurrence at the base of the Guadalupe section is somewhat surprising on one hypothesis, and the abundance of the strongly plicated Enteletes hemiplicatus in the "Coal Measures " of the Mississippi Valley almost equally difficult of explanation on the other.

With the exception of a single specimen of Enteletes from the Delaware Mountain formation, this is the only occurrence of any branch of the family known up to the present time in the Guadalupe section, and the complete absence of Enteletes, or, in fact, of orthoids of any sort, in the higher beds is noteworthy.

In addition to specimens from the black limestones, on which the foregoing description is based, I have subsumed under this title a dorsal and a ventral valve belonging to the same individual obtained in the Diablo Mountains. These specimens, which have unfortunately been lost since the drawings were made, are figured on Pl. XXVI of the present volume. Save for an indistinct fold and sinus they are entirely without plications, and as they possess the internal structure of Orthotichia as well as of Enteletes, I originally referred them to the former genus. I do not, however, see that it would be possible to distinguish them from small examples of the present species or some related form of Enteletes. Compared specifically with the form under discussion there are few differences by which they may be distinguished. The liration of the shell from the Diablo Mountains is of the same general character, but much fainter than that of well-preserved specimens from the black limestone. This lack of distinctness, however, I suspect to be due in some measure to siliceous replacement. The dorsal valve is perceptibly less convex, but this as a rule must be regarded as a character of inferior importance. Dorsal valves from the black limestone, moreover, show a distinct cardinal area, and though the structure is absent in the delicate silicified specimens from the Diablo Mountains, it may well be supposed to have been broken away. These silicified specimens, which, as above noted, are less tumid than the others, appear to be related to a species described by Waagen as Orthis derbyi,a and more especially to a variety of this described by Schellwien, with a change of generic designation, as Enteletes derbyi var. demissa.b From the latter it can probably be distinguished by being more nearly equivalve and by the: faintness of its striation. No specimens of the European form, however, are available for comparison. The specimens from the Diablo Mountains occur in association with several species of Enteletes, but any one of these, at the size which they have reached, would have developed lateral plications, no sign of which appears on the specimens under consideration.

---

aSalt Range fossils: Mem. Geol. Survey India, Pal. Indica, ser. 13, vol. 1, p. 565, pl. 56, figs. 2, 5, 6.

bAbbandl. K-k. geol. Reichsanstalt, vol. 16, part 1, p. 8, pl. 1, figs. 4-7.

---

The specimens from the black limestone also present considerable external resemblance to Orthis derbyi, but like those from the Diablo Mountains they show within three high, thin, subparallel septa in the ventral valve, different from the low curved ridges surrounding the muscular imprints in the Indian species. Because of these structures the Guadalupian fossils have been placed in a different genus, namely, in Enteletes.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (stations 2920 and 2967). Delaware Mountain formation, Diablo Mountains, Texas, as reported (station 3764).

ENTELETES sp. d.

Pl. XXI, fig. 23.

From the yellow sandstones of the Delaware Mountain formation has come to hand a single dorsal (?) valve of Enteletes whose preservation is imperfect, so that many characters can not be made out with certainty. It is strongly inflated and has a subcircular shape. The length is 20 mm. and the width, if complete, probably not less than 25 mm. There are certainly seven plications, and possibly one or two more, and a considerable area at the sides is unplicated. The plications are not so large or so sharp as in E. hemiplicatus, but the liration appears to be of about the same size and character. It is impossible to tell from the present specimen whether it belongs to the dorsisinuati or to the ventrisinuati.

This form is undoubtedly distinct from our common E. hemiplicatus, and it is also clearly distinct from E. globosus and E. angulatus. It has less coarse plication than E. dumblei, as well as different striation, and probably can be distinguished from any of the three unnamed forms already described in this report. It most closely resembles the form denominated Enteletes sp. b, but in configuration is not dissimilar to Enteletes sp. a, differing, however, in having the liræ more distinct and equal, seemingly without the prominent spiniferous ones. Of foreign species, probably E. elegans Gemmellaro is most closely allied.

I have also provisionally placed with this species some fragmentary crushed material from station 3501. These specimens, if complete, would probably be of considerable dimensions. Two examples, showing only the apical portion, have a length of 16 mm. The liræ are moderately fine, rounded, and among the ordinary ones portions of the surface show others of greater prominence, which were doubtless the bases of small spines. Plications on these fragments are not very distinct, but small, and apparently rather numerous. On the more mature areas the plications were correspondingly more pronounced. This form appears to belong to the same group as Enteletes sp. a, Enteletes sp. b, and Enteletes sp. d. In the unequal character of its liræ it suggests the type constituting Enteletes sp. a.

The general character of this form is especially suggestive of the shell from the Delaware Mountain formation on which Enteletes sp. d, is founded, the chief difference being in that the plications in the latter extend farther toward the beak. The shell from station 3501, however, has apparently been considerably worn. In the faint development of plications, at least over the umbonal area, these shells suggest the form from the black limestone here designated Enteletes sp. c. The plications are somewhat more distinct, and are smaller than I judge they would be in the latter species.

A very young shell from station 3763, the horizon of which is probably equivalent to some portion of the Delaware Mountain formation, has also been placed in this group. This specimen, which is of course entirely without plications, might be either a Schizophoria, an Orthotichia, or an Enteletes, but from the strength of the three septal plates I judge that it belongs to the latter genus. Its specific relations are of course largely speculative.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2919). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763?). Delaware Mountain formation, southern Delaware Mountains, Texas (station 3501).

Family PENTAMERIDÆ McCoy.

In the Carboniferous the family of the Pentameridæ is restricted to the single genus Camarophoria, and in the Guadalupian the genus is represented by but a single species. Camarophoria is much more abundant at a lower horizon (in the Hueconian), and it is more abundant in the Salt Range, from which Waagen recognized no less than five species. The Guadalupian form probably belongs to what Waagen calls the group of C. crumena, in which three Salt Range species are included. The group of C. rhomboidea, comprising two Salt Range forms, is unknown in the Guadalupian.

Diener recognized three species of Camarophoria in the Chitichun fauna No. 1, of which one, identified as Camarophoria sp. indet. aff. C. crumena, is unlike the Guadalupian form. The two others, however, belong to the same general type. As an immature stage of C. crumena, Diener figures a little shell which, if the same genus, certainly appears to belong to a different species. It is altogether different from the Guadalupian species of Camarophoria, and looks as if it might be a Pugnax of the general type of P. utah, etc. In his second paper on this fauna Diener cites, in addition to C. purdoni var. gigantea, which is related to the Guadalupian species C. venusta, Camarophoria cf. semiplicata, possibly the same as the small specimen commented on above, and C. globulina. The two latter are of an altogether different type from C. venusta.

In his paper describing the Anthracolithic fossils from Kashmir and Spiti this author mentions only Camarophoria cf. purdoni, related in some degree to the Guadalupian species. His later discussion of this fauna mentions only Camarophoria n. sp. from the lower beds, a form not related to C. venusta, but makes no record of the genus in the upper beds. Diener cites C. purdoni from Malla Sangcha also.

Like the rhynchonelloids, the Pentameridæ appear to be conspicuously rare in the Carboniferous faunas of China and adjacent areas. Kayser records none from the Lo Ping locality, and Loczy cites only Camarophoria purdoni from the Lantsankiang Valley, among the different Chinese Carboniferous faunas which he describes. This form is related to P. venusta. Camarophoria crumena is cited by Beyrich from Timor, the form illustrated being of the same general shape as C. venusta. In his paper on fossils from Timor and Rotti Rothpletz cites Camarophoria pinguis, which, to judge from his figures, is somewhat remotely related to the Guadalupian species.

Tschernyschew cites C. crumena and C. margaritovi from Vladivostok. C. margaritovi is described as a new species and is less closely allied to C. venusta than is, presumably, C. crumena, which is not figured.

In the Russian section Camarophoria is represented in the Moskovian stage by two species, which Trautschold identifies as C. crumena and C. plicata. C. crumena, which alone is figured, does not resemble the Guadalupian form very closely. In the Gschelian this type shows a remarkable differentiation, Tschernyschew recording no less than 15 species. The 5 species included in the group of C. crumena show individually rather wide variation. Some of the forms included there appear to be distinctly related to C. venusta, others very slightly so. Some of them have much more the configuration of certain of the Guadalupian species of Pugnax. The group of C. applanata, which consists only of the species of that name, is also related to C. venusta, but the group of C. isorhynchus, with three species, the group of C. rhomboidea, with four, and the group of C. sella, with two species, have little to do with the single known Guadalupian form. These groups show wide variations of configuration, some of which are remarkably dissimilar to C. venusta. C. superstes is suggestive, in configuration at least, of Pugnax bisinuata, and C. globulina of Pugnax bidentata, but, so far as known, they are essentially different.

Stuckenberg records from the Gschelian C. plica, C. purdoni, C. sella, C. biplicata, and C. triplicata. Some of these forms, as already remarked, such as C. plica and C. sella, are only very remotely related to C. venusta. In some of the others the relationship is more obvious. Nikitin also records C. purdoni from this horizon, and his figures show a form in a general way comparable to the Guadalupian species.

From the Artinsk Tschernyschew cites only C. plica, a species belonging to a different group from C. venusta. Stuckenberg recognized C. duplicata and C. purdoni from this horizon, two species which if not very closely allied to C. venusta are more so than C. plica. One of the same forms (C. purdoni) is recorded from the Kungurstufe.

Netschajew found six species of Camarophoria among his Permian fossils from eastern Russia. C. superstes, C. waageni, and C. globulina are not closely allied to the Guadalupian Camarophoria venusta, but C. humbletonensis, C. purdoni, and Camarophoria cf. schlotheimi probably belong to the same group. The species last named is not figured, however, and Netschajew's specimens of the two others were so poor that a trustworthy conclusion is impossible. Tschernyschew found in the Permian of the government of Kostroma only the species Camarophoria superstes, a form altogether dissimilar superficially to C. venusta and having much the configuration of Pugnax bisinuata. De Verneuil in his classic work on the Russian Permian recognized only C. schlotheimi and C. superstes, two species which are unlike the Guadalupian representatives of the genus. In configuration they suggest, respectively, Pugnax bisinuata and P. swallowiana, as I have already had occasion to remark, although the fact is apparently without significance save as an instance of parallel development.

In a general way there is more resemblance shown between the Guadalupian representatives of Camarophoria and those of the Russian Permian than between the Guadalupian Camarophorias and those of the Russian Gschel fauna. This is perhaps due rather to negative than to positive elements—to the diminished occurrence of the group in the Permian and the presence of fewer forms which are very unlike the single Guadalupian species. The genus is better represented, both in abundance of individuals and differentiation into species, in the upper beds of the Hueco formation, which underlies the Guadalupian.

Camarophoria seems to be absent in the Armenian fauna described by Abich and reviewed by Arthaber, but it appears among the fossils from Balia Maaden, in Asia Minor, which Enderle described. Enderle identified his fossils as C. globulina, and his figures do not indicate a close relationship with C. venusta.

Gemmellaro distinguished five species among the Silician Camarophorias, all of which belong to different groups from C. venusta. Most of them have the superficial aspect of our American Pugnaces (C. affinis of Pugnax rockymontana and some of the others of P. utah, P. swallowiana, etc.). C. acuminata, however, is more like the Guadalupian species referred to Rhynchonella.

In the fauna of the Carnic Fusulina limestone Schellwien found three species of Camarophoria which he describes as C. alpina, C. sanctispiritus, and C. latissima. There is no question of identity with these species, but the Guadalupian form is of the same general type. This author recognized three species also in the fauna of the Trogkofelschichten. In contrast to those recorded in his earlier report, the three Trogkofel species only remotely resemble the Guadalupian C. venusta.

The Dyas of Germany contains a considerable series of forms belonging to this genus all of which are included by Geinitz under the title of C. schlotheimi. Most of them are altogether different in general appearance from C. venusta, but a few are less remotely related.

The Permian of England furnishes a greater variety and King recognizes three species. By far the most closely allied to C. venusta is the English form identified as C. multiplicata. The relationship is less distinct in the case of C. schlotheimi and remote in the case of C. globulina.

Toula cited C. crumena (without figures) from the south point of Spitzbergen, and a little later reported the same species from the Hornsund, his figures in the latter case showing a form belonging to a different group from the Guadalupian species.

The only reference to this genus in the several works consulted which deal with the Carboniferous of South America is found in Derby's account of Brazilian fossils. He there mentions fragments of a small species of Camarophoria too imperfect for description.

The absence of Camarophorias in the typical Pennsylvanian faunas constitutes one of their marked peculiarities. The genus is somewhat sparingly represented in the Mississippian, but so far as known does not in the eastern region range into the upper beds. In the West we have Camarophoria thera, which is different from C. venusta, and Camarophoria bisinuata, described by Shumard from the present fauna, which has proved to belong to an entirely different group. In the Hueco formation the genus is better represented than in any of the horizons and areas mentioned. In California and Alaska at horizons tentatively correlated with the Russian Gschelstufe, the genus is also found in relative abundance.

Genus CAMAROPHORIA King.

CAMAROPHORIA VENUSTA n. sp.

Pl. XXXI, figs. 6 to 6c.

Shell rather large, subtriangular. Length and breadth nearly equal. Ventral valve shallow. Beak elongate, pointed, slightly incurved, and apparently without deltidial plates. Sinus moderately deep and well defined; distinguishable about halfway back from the front margin. There are five plications in the sinus and about six on either side.

Dorsal valve strongly convex, fold moderately high and well defined, surmounted by six plications, in addition to which there are about six lateral ones on each side.

The plications are rather slender and strongly rounded or subangular. The lateral ones are lower, rounder, and less distinct than those on the fold and sinus.

I have identified this species, the type of which was found in the Glass Mountains, somewhat doubtfully in an imperfect specimen from the white limestone of the Guadalupe Mountains. The chief difference to be seen at present is that the Guadalupe specimen has somewhat coarser and therefore fewer plications. This example shows a fine large spondylium in the dorsal valve. Aside from this the generic position is not well determined and rests chiefly on the general expression and apparent relation to other species belonging to Camarophoria. C. venusta is evidently the representative of C. purdoni and C. humbletonensis of the Salt Range fauna. The last-named species, it will be remembered, is found also in the Permian strata of England.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Family RHYNCHONELLIDÆ Gray.

According to my present knowledge the Rhynchonellidæ of the Guadalupian fauna are represented by some 16 species, which have been assembled under two generic titles. So far as known, however, the two groups differ more in configuration than in structure. In both the ventral valve is provided with a pair of more or less well-developed dental plates, while the dorsal has a low but distinct median septum. As to configuration, however, the shells placed with Pugnax have a few large plications which reach only part way to the beak, those on the sides tending to be obsolete. The fold and sinus are well developed and distinctly defined, and occupied by more pronounced plications than the lateral ones. The forms referred to Rhynchonella?, on the other hand, have numerous relatively fine plications, which reach quite to the beak and often increase by bifurcation. The fold and sinus are well developed, but by reason of the occurrence of plications on their sides are not always sharply defined.

Even among the shells referred to Pugnax two rather distinct groups are found, one represented typically by P. bisulcata, with very gibbous growth, subcircular outline, and rounded, almost obsolete plications, the other with triangular or pentagonal outline and ribs sharp and strong, especially on the fold, though not persistent.

Waagen distributed his Salt Range rhynchonelloids among the genera Terebratuloidea (4 species), Uncinulus (3 species), and Rhynchonella (3 species). It would thus appear that the only generic group which is shared by the Guadalupian fauna is Rhynchonella, the Salt Range fauna being without Pugnax and the Guadalupian without Terebratuloidea and Uncinulus; nor can it be said that the specific development of Rhynchonella in the two faunas shows much parallelism.

In his first paper on the Chitichun fauna No. 1, Diener recognizes only one species of rhynchonelloid, which he identifies as Uncinulus timorensis. Nothing yet known in the Guadalupian appears to be related to this. His later paper on this fauna cites Terebratulidea cf. depressa and Uncinulus jabiensis, neither of which from our present knowledge can be correlated with any Guadalupian species.

Diener did not find any representatives of this family among the collections on which was based his paper describing the Anthracolithic fossils from Kashmir and Spiti, but Davidson described three species from Kashmir under the titles Rhynchonella pleurodon var. davreuxiana, R. barusiensis, and R. kashmiriensis. The latter may be compared with the Guadalupian species Pugnax pinguis and the form which I have identified as P. osagensis, but the two other species do not possess related types in the Guadalupian. In a later paper on the Spiti Anthracolithic fossils Diener cites Rhynchonella confinensis and Rhynchonella cf. wynnei from the lower beds, but no species from the upper. R. confinensis more closely resembles some of our Mississippian species of Camarophoria than any Guadalupian form, and Rhynchonella cf. wynnei also seems to be without any closely related Guadalupian species. The same author among his fossils from Malla Sangcha recognized Uncinulus timorensis, U. jabiensis, and Rhynchonella sp. indet. ex aff. R. hofmanni, none of which, however, appears to be allied to Guadalupian species.

From Turkestan Romanowsky recognized four rhynchonelloids, which he distinguished as Rhynchonella daleidensis Roem, R. turanica n. sp., and two species undetermined. One of the latter suggests by its form a Pugnax related to P. swallowiana and P. osagensis, but the three other species appear to be unrelated to anything in the Guadalupian.

A brief consideration of the Carboniferous faunas of the Salt Range and of the Himalaya, and of the probably older (in part, at least) faunas of Turkestan, shows that they have no very close resemblance to the Guadalupian fauna in point of the Rhynchonellidæ.

Kayser found no representatives of the Rhynchonellidæ among his fossils from Lo Ping, but Loczy cites Uncinulus timorensis from the Lantsankiang Valley. The general appearance of this form, as presented by his figures, is not unlike some of the forms which I have referred to Rhynchonella (such as R. indentata and R. longæva), but presumably they are not essentially related.

Roemer cites Rhynchonella cf. R. pleurodon from the west coast of Sumatra, and Beyrich describes Rhynchonella timorensis from Timor. The last-named species in its configuration strongly suggests a relationship with the form from the black limestone, which I have described as Pugnax? nitida. Subsequent authors, however, have placed Beyrich's species in the genus Uncinulus, to which group, even as it was defined by Waagen, the Guadalupian shell probably does not belong. Furthermore, Rothpletz's figures of this species in his paper on the fauna of Timor and Rotti do not so much resemble P. nitida. This author cites from Timor and Rotti, in addition to Uncinulus timorensis, a species which he describes as Rhynchonella wichmanni. It appears to be related to R. indentata and R. longæva.

The Rhynchonellidæ of this geographical group of faunas also do not show much affinity with the Guadalupian representatives of the family. Rhynchonelloids seem in fact to be rather rare and to be lacking in many of the characteristic Guadalupian types. An American paleontologist would perhaps be especially struck by the absence in these faunas and in those of India of any representatives of the type so characteristic of the Pennsylvanian of North America, which is commonly identified as Pugnax utah.

Only two Rhynchonellas are mentioned by De Koninck in his account of the Carboniferous faunas of New South Wales, One of these seems to belong in the earlier horizon and the other has been shown to be a Dielasma.

The only rhynchonelloids which are recorded by Etheridge from the "Permo-Carboniferous" of Queensland and New Guinea are Rhynchonella pleurodon and Rhynchonella sp. indet., the latter not figured.

In comparing the Guadalupian Rhynchonellidæ with those of the Russian section it has seemed best to disregard such references as are unaccompanied with descriptions, and especially with figures. This course was judged advisable, not only because owing to the great variability of these shells very different types have sometimes passed as the same species, but also because there is very little uniformity, even in the generic terminology.

Trautschold in his monograph on the Moskovian fauna recognized only one species of rhynchonelloid, if we may suppose the two forms placed with Camarophoria to be really members of the Pentameridæ. The species identified as Rhynchonella pleurodon is presumably a Pugnax, and, if so, appears to be more nearly allied to P. shumardiana than to any other Guadalupian representative of this group.

In the Gschelian fauna Tschernyschew refers his rhynchonelloids to the genera Rhynchonella (3 species), Terebratuloidea (2 species), Pugnax (8 species), Uncinulus (1 species), and Rhynchopora (3 species). This fauna contains many forms that are not found in the Guadalupian—in the present case representatives of the genera Terebratuloidea, Uncinulus, and Rhynchopora. Some of Tschernyschew's figures of Uncinulus wangenheimi suggest the forms here placed with Rhynchonella, but I am unable to refer the Guadalupian shells to Uncinulus. The three species which Tschernyschew refers to Rhynchonella are not very similar to those which I have placed with that genus, appearing in fact in one case to resemble superficially some of the Guadalupian Pugnaces. R. granulum at least suggests Pugnax? pusilla of the present work.

A number of Tschernyschew's species of Pugnax appear to be related to the Guadalupian forms referred to P. swallowiana and P. osagensis. Such are P. osagensis, P. swallowiana, P. kayseri, Pugnax sp., and Pugnax n. sp. P. connivens is in part comparable to P. bidentata, but it almost appears as if two species were figured under that name. P. keyserlingi and P. granum have no known representatives in the Guadalupian fauna, while, on the other hand, the group of P.? bisulcata, together with P. nitida and probably also P. shumardiana, P. elegans, and P. pinguis, as well as the Guadalupian Rhynchonellas, have no corresponding forms in the Russian fauna.

Nikitin cites among the Rhynchonellidæ only Rhynchopora nikitini from the Gschelian horizon and Stuckenberg only Rhynchopora nikitini and R. variabilis.

From the Artinsk Stuckenberg records Rhynchopora nikitini and Rhynchonella sp. (not figured), and from the Kungurstufe Rhynchopora variabilis. Tschernyschew cites from the Artinsk only Rhynchopora nikitini and Rhynchonella hofmanni (not figured). The only Rhynchonellas cited by Krotow in his paper on the Artinsk fauna are Rhynchonella pugnus, Rhynchopora pleurodon, and Rhynchopora geinitziana. None of the identifications is figured and it would be hazardous to do more than again recall that two at least of the species belong to a genus which is absent from the Guadalupian fauna.

In the Permian of the Government of Kostroma the only rhynchonelloid found by Tschernyschew is identified as Rhynchopora geinitziana, and in the Permian of eastern Russia Rhynchopora nikitini. The same species are recorded by Netschajew.

Considered as a whole the Guadalupian rhynchonelloids are not closely related to those of the Russian section, but are nearest to those of the Gschelian zone. In the Russian Permian the only genus representing the Rhynchonellidæ appears to be Rhynchopora, a type whose absence from the Guadalupian fauna is rather remarkable. The Guadalupian species called Pugnax pinguis suggests Rhynchopora in its configuration, but appears to have an impunctate shell. Rhynchopora and Rhynchonella, one species of which (Rhynchonella hofmanni) has no corresponding Guadalupian form, seem to constitute the entire rhynchonelloid development of the Artinsk.

Even in the fauna of the Gschelstufe the resemblance to the Guadalupian consists principally in the presence of certain types of Pugnax related to P. osagensis, P. swallowiana, etc. At the same time it contains generic groups (Uncinulus, Terebratuloidea, and Rhynchopora) not found in the Guadalupian, besides types of Rhynchonella and Pugnax which are not known there. On the other hand, the Rhynchonellas of the Guadalupian are not closely related to the Gschelian representatives of that genus, while several types of Pugnax found here, such as P. bisulcata, P. nitida, and P. pinguis, are not known in the Russian fauna.

Among the fossils from Djoulfa, in Armenia, Abich recognized Rhynchonella pleurodon, and figured two somewhat different specimens under that title. One of them suggests the Guadalupian species which I have described as Pugnax pinguis. The other also has somewhat the same resemblance. The latter, but not the former, was subsequently included by Arthaber in his work on this fauna, in the species Uncinulus wichmanni. Arthaber's figures, however, remind one somewhat of the Guadalupian Pugnax nitida. Arthaber also cites Uncinulus jabiensis, figuring two specimens which certainly do not appear to be the same species. Superficially the larger of the two specimens so identified somewhat resembles Pugnax shumardiana and the smaller P.? pusilla.

The only rhynchonelloid from Asia Minor which Enderle recognized in his paper on the fauna of Balia Maaden is cited by him as Rhynchonella cf. triplex McCoy. In configuration it is suggestive of the Pugnaces, and may be compared, in this particular, at least, especially to P. elegans.

Gemmellaro referred his rhynchonelloids of the Fusulina limestone of Palermo to the genera Rhynchonella, Uncinulus, and Terebratuloidea, the former comprising 7, the second 3, and the latter 1 species. Gemmellaro's Rhynchonellas seem to belong to two groups, one of which is certainly very suggestive of the forms which I have placed with Pugnax, and the other of those which I also have called Rhynchonella, His Rhynchonella negrii, R. sosiensis, and R. adrianensis appear to be related to the Guadalupian shells cited as Pugnax swallowiana, P. bidentata, and P. osagensis. Rhynchonella withei resembles, though certainly rather remotely, P. nitida. Rhynchonella salinasi, however, rather suggests Rhynchonella indentata of the Guadalupian, but Rhynchonella carapezzæ probably has no corresponding form. Rhynchonella acuminata, also, has no Guadalupian species at all related, so far as I am aware, nor does the Sicilian fauna contain species which are really close to P. bisulcata, P. bisulcata var. seminuloides, and P. nitida, or possibly also to P. pinguis and P. shumardiana.

None of the Guadalupian rhynchonelloids can, in my opinion, be properly referred to the genus Uncinulus, but some of the Sicilian species of Uncinulus show a superficial resemblance to some of the Guadalupian species of Rhynchonella and Pugnax—e. g., Uncinulus amor to Pugnax? nitida; Uncinulus velifer and Uncinulus siculus to Rhynchonella indentata, R. longæva, etc. This is not true of the Sicilian species of Terebratuloidea, for there is nothing in the Guadalupian similar to it.

In his paper on the fauna of the Carnic Fusulina limestone Schellwien recognizes two species of rhynchonelloids which he refers to the genus Rhynchonella itself. One of these, R. grandirostris, has the general expression of the genus Pugnax and may be related to the Guadalupian types referred to P. swallowiana and P. osagensis. The other species is unlike anything at present known from the Guadalupian beds.

The rhynchonelloids of the Trogkofelschichten are referred by Schellwien to the genera Uncinulus, Rhynchonella, Pugnax, and Terebratuloidea. The single representative of Uncinulus in this fauna, U. velifer, suggests by its configuration the Guadalupian shells referred to Rhynchonella, Schellwien's Rhynchonella confinensis, as already remarked, has no species in the Guadalupian at all related. The two species representing the group of Rhynchonella pleurodon—R. wynnei and Rhynchonella aff. sosiensis—have somewhat the expression of the Guadalupian Pugnax osagensis and P. pinguis. Only one Trogkofel form has been referred to Pugnax. It is related to the Guadalupian species of the type of P. swallowiana, but it is especially semblable to R. shumardiana. In configuration, at all events, the two Alpine species of Terebratuloidea are less unlike the Guadalupian rhynchonelloids than most of the foreign representatives of that genus. They have somewhat the expression of the Guadalupian Pugnax swallowiana, P. osagensis, etc.

The relationship between Schellwien's fauna and that from the Guadalupe Mountains seems to me rather slight. One can not but be struck by the differentiation in the American fauna of shells of the type of P. osagensis, and their relative scarcity in that of the Carnic Alps. The latter has also nothing to compare with P. bisulcata, P. nitida, etc., and but little apparently to be correlated with the Guadalupian Rhynchonellas. On the other hand, the Guadalupian fauna is without Terebratuloidea, Uncinulus, etc. The Guadalupian rhynchonelloids appear to me more closely allied to those of Palermo, and yet the correspondence in this group of shells is not particularly striking.

Gortani cites the Pennsylvanian species Rhynchonella osagensis from the Carnic Alps. It is not figured, but is presumably related to Pugnax osagensis and P. swallowiana of the present fauna.

In the German Dyas the only rhynchonelloid recorded by Geinitz is referred to Rhynchopora geinitziana Vern. In the reduction of the rhynchonelloid representation practically to this genus the Dyas resembles the typical Russian Permian, while in the English Permian, so far as King's monograph is conclusive, the entire group is wanting. In this respect, as in certain others, the Guadalupian fauna presents a distinct point of difference from either of these Permian facies.

Among the fragmentary faunas described from Spitzbergen the only mention of representatives of this family which I have come upon are Rhynchonella cf. pleurodon from Axel Island and Rhynchonella sp. indet. from the cape between the two arms of North Fjord. These two forms are closely related, perhaps the same species, but it is doubtful if they have any corresponding Guadalupian types. Toula cites the same species, but without figures, from Nova Zembla.

Stache found Rhynchonella cf. trilatera and Rhynchonella aff. carringtonensis at two stations in the West Sahara, but the associated faunas are probably older, and the forms themselves, at all events, have little in common with any Guadalupian species.

Salter cites from Bolivia a species of Rhynchonella which he compares to R. pleurodon. It is probably a Pugnax, and at all events has a configuration which much resembles that of P. shumardiana, P. swallowiana, and P. osagensis. Under the title of Rhynchonella pleurodon Toula also cites a similar species from Bolivia. Its relations appear to be with the same group of species, perhaps with P. shumardiana.

The only rhynchonelloid found by Derby among his Brazilian species was described as new, under the title Rhynchonella pipira. From certain characters set down in his description it seems likely that this form belongs to the group of shells for which Waagen and a number of European authors have employed the term Uncinulus. This group appears to be unrepresented in the Guadalupian fauna. In a general way Rhynchonella pipira might perhaps be compared to Pugnax pinguis of the Guadalupian, and perhaps to other forms, but probably none of them is closely related to it.

If we eliminate certain western species, especially those which have been described from the fauna under consideration, and a few forms from the Mississippi Valley described without figures and consequently pretermitted except in catalogues and bibliographies, the upper Carboniferous or Pennsylvanian rhynchonelloids of North America, so far as known, resolve themselves into four groups representing two generic types. Rhynchopora illinoisensis is rather rare, and Pugnax rockymontana is also seldom found. The only species which can be looked for at this horizon to appear frequently and in abundance is the form, or group of forms, for which the name Pugnax utah Marcou is most often used. A related species is the form which Hall and Clarke have incorrectly identified as Pugnax swallowiana. The genus Rhynchopora is unknown in the Guadalupian fauna, and Pugnax rockymontana, or any form like it, does not occur there. Species related to P. osagensis and P. swallowiana, however, form nearly the most abundant rhynchonelloid type. On the other hand, the group of P. osagensis presents modifications not found in the Pennsylvanian fauna, such as appear in the species P. bidentata, P. shumardiana, P. elegans, P. pusilla, and P. pinguis, nor is anything at all resembling P. bisulcata, P bisulcata var. seminuloides, or Rhynchonella indentata, R. longæva, etc., known from the Pennsylvanian. Except for a few species of very wide dispersion, of the general type of P. utah (P. osagensis), therefore, the Guadalupian Rhynchonellidæ and those of the Pennsylvanian have nothing in common and, a great deal which is peculiar to each.

Genus PUGNAX Hall and Clarke.

In all, 12 Guadalupian species have been referred to this genus, and they comprise two rather distinct types. One of these is exemplified by our common Pugnax osagensis (P. utah auctorum). It has a few strong angular plications, which as a rule extend only about halfway back from the margin. In the other group the plications are fainter and still more marginal, practically obsolete on the sides and obsolescent on the fold and sinus. In this latter group are included only Pugnax bisulcata, P. bisulcata var. seminuloides, and P. bisulcata var. gratiosa. The eight remaining species belong to the group of P. utah, but even among these some sort of subdivision can be effected. P. pinguis and P. pusilla are distinguished from the six other species by the more rounded and persistent nature of the ribs, but the form identified as P. osagensis is intermediate in some degree and these two groups are not as distinct from one another as is the group of P. bisulcata from either of them.

Structurally these groups, so far as I have ascertained, present but slight differences from one another or from the series of species which I have placed with the genus Rhynchonella sensu stricto. In all these forms there are present two well-developed dental plates, and in the dorsal valve a low but distinct median septum. The upper Carboniferous Pugnaces, like P. osagensis, have a better developed dorsal septum than is generally credited to them, so that in this particular, as well as in configuration, no disagreement is evinced with the Pennsylvanian forms. One structural difference seems to exist between the shells of the group of P. osagensis and those of P. bisulcata and its allies, namely, the less extensively developed and differently shaped hinge plate of the latter; and I am not sure but that this circumstance, taken in connection with the difference in configuration, would have made it warrantable to assign them to different genera.

The shells referred to Rhynchonella have, so far as ascertained, similar internal structures to those placed with Pugnax. The singular feature observed in one example of Rhynchonella of a perforated hinge plate may serve as an index of other differences that will be manifest when the two types are perfectly known. Externally the Guadalupian Rhynchonellas are distinguished by having numerous fine, sometimes bifurcating plications, which extend quite to the beak, instead of a few large incomplete ones. These differences, although of configuration alone, are sufficiently marked to render the relationship of species entirely unambiguous.

The Guadalupian shells referred to Pugnax seem to be distinct from many of the different rhynchonelloid genera which have been discriminated in the Carboniferous faunas of Asia and Europe. They are without the punctate shell structure of Rhynchopora. They have less persistent plications than Terebratuloidea, lack the truncated beak and round foramen of that genus, and possess different internal structures, since they have a median septum and dental plates.

From Uncinulus they differ in having much fewer and larger ribs, which are not furrowed along the top. Internally the structure appears to be about the same, save that the Pugnaces do not possess the thickened ventral shell near the septum far from the apex of the ventral valve.

Rhynchonella as usually identified comprises shells having a similar internal organization but finer, more numerous, and more persistent plications. There is not entire uniformity among authors in the usage of these terms, however, especially in that of Rhynchonella. Gemmellaro, for instance, includes in that genus forms which seem to me to have the general expression of Pugnax.

PUGNAX? BISULCATA Shumard.

Pl. XXI, figs. 11 to 12.

1858. Camarophoria (?) bisulcata. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 296 (date of volume, 1860).

Dark Permian limestone: Guadalupe Mountains; conglomerate at the mouth of Delaware Creek.

1859. Camarophoria bisulcata. Shumard, idem, p. 394, pl. 11, figs. 2a to 2d.

Dark and white [Permian] limestone: Guadalupe Mountains; conglomerate at the mouth of Delaware Creek.

Shell variable, outline varying from nearly circular to subpentagonal, with angles obtusely rounded, sometimes very gibbous and sometimes moderately so, usually a little transverse, sides always rounded, front sinuate; shell structure fibrous. Ventral or receiving valve very depressed, gently convex, greatest convexity near the beak; cardinal margins forming an obtuse angle; mesial sinus broad at the front, scarcely reaching the middle of the valve, shallow or rather deep, perfectly smooth or bearing from two to five obscure, rounded ribs; tongue of sinus moderately produced, broadly truncate at extremity, and curved upward, sometimes at nearly a right angle with the general surface of the valve; beak imperforate, pointed, incurved nearly in contact with the opposite valve. Dorsal valve strongly rounded in most specimens, much more gibbous than the opposite valve, marked with a broad, shallow depression or false sinus extending from beak to front, which is bounded on either side by a ridge very obscure on the rostral half of the shell, but forming together a broad mesial fold toward the front, which is smooth, or marked with two or more slightly prominent plications; beak rounded, obtuse, extremity usually hidden by the beak of the opposite valve.

Dimensions of an average specimen: Length, 0.58; width, 0.63; height, 0.35.

Resembles Terebratula superstes Verneuil, from which it is distinguished by the greater convexity of the dorsal valve and its more flattened ventral valve.

Restricted to the dark limestone of Permian age at the base of the white limestone of the Guadalupe Mountains; found also abundantly in the conglomerate at the mouth of Delaware Creek, New Mexico.a

---

aTrans. Acad. Sci. St. Louis, vol. 1, 183-1860, p. 296.

---

Shumard's description, which is quoted in full above, includes three varieties, two of which I shall describe as new, under the names Pugnax? bisulcata var. seminuloides, and P.? bisulcata var. gratiosa. As indicated by our collections, the variety seminuloides is by far the more abundant; but because Shumard figured only the plicated form I retain for it the name originally applied to all three. So similar are the three varieties in general appearance, however, that it becomes necessary to modify Shumard's description but little, in spite of the shells which have been with drawn from it. I propose to restrict P.? bisulcata to shells having the fold marked by four or five narrow plications, and to use the varietal names seminuloides and gratiosa for those in which the fold is unplicated and marked by two or three broad plications, respectively.

The shape, of P.? bisulcata is usually subcircular, and somewhat transverse. The ventral valve is rather shallow, the dorsal frequently very gibbous. The ventral beak is small and strongly incurved, not erect, as represented in Shumard's figures. The fold is high, but its limits are not sharply defined, and it can be observed only a short distance back from the front. From three to five plications have been noted on the fold, and a corresponding number on the sinus. I scarcely understand Shumard's description of the dorsal valve as having a shallow false sinus. My specimens, which are, unfortunately, somewhat crushed, show no peculiarity of this sort. The sides are apparently smooth, although very rarely a few evanescent ribs occur on either side near the fold.

Shumard remarks on the occurrence: "This species was found very abundantly in the dark limestone beneath the white limestone of the Guadalupe Mountains and very sparingly in the white limestone. It is also quite common in the conglomerate at the mouth of Delaware Creek, Texas."b My material is entirely from the "dark limestone," though I have a single specimen of the variety seminuloides from the white limestone. I shall discuss the generic position of the species when considering this variety, of which my material is more abundant.

---

bIdem, p. 394.

---

The forms assembled under this title are, by applying the same fundamentum divisionis which is used in other rhynchonelloids, susceptible of division into a number of species or varieties. Some have lateral plications and some none, while a good deal of variation (shown by the illustrations to some extent) is manifested in the number, size, and sharpness of the plications. For the present, however, it seems best, while not ignoring these differences, to make them the basis of varietal names only.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PUGNAX? BISULCATA var. GRATIOSA n. var.

Pl. XXI, figs. 10 to 10c.

This variety is associated with typical P.? bisulcata, and resembles it closely in most particulars. The chief difference, as in the variety seminuloides, concerns the plications on the fold and sinus. The present variety, as based on the typical specimen, has three obscure, broad plications on the fold; P.? bisulcata has four or five narrow ones; and P.? bisulcata var. seminuloides has none at all.

That Shumard included the present variety in the description of his Rhynchonella bisulcata there can be little doubt, but his figures seem intended to depict the variety of which the original of my fig. 12 of Pl. XXI is a representative specimen. I think that the latter may be regarded as the typical variety, and that the name bisulcata may be restricted to it. It seems hardly desirable, however, to include the form represented by fig. 10 without any distinction under Shumard's species, and I have accordingly discriminated it as a variety.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PUGNAX? BISULCATA var. SEMINULOIDES n. var.

Pl. XXI, figs. 13 to 16; Pl. XXIX, fig. 9?.

Shumard, as already remarked, includes this form with Pugnax? bisulcata. When well-marked examples are kept in view it would seem that surely here must be two distinct species. On the other hand, the plications of P.? bisulcata become so faint in some specimens that the probability of complete intergradation must be admitted. Nevertheless, one is seldom at a loss, if a specimen is at all well preserved, in assigning it to one group or the other. On this account, while refraining from giving both forms full specific rank, as I was at first disposed, it seems to me best to distinguish the unplicated one as a distinct variety. So close is the resemblance in general configuration of the variety seminuloides to Shumard's species that no separate description is necessary, the discriminating feature being the absence of plication on the fold or, indeed, anywhere on the shell. There seems to be a tendency, likewise, for the fold to be broader, lower, and with less sharply defined limits, though this is not invariably the case. A gentle mesial depression is sometimes present, which has perhaps been described by Shumard as a false sinus in the dorsal valve of this species. I have not observed it in all nor even in a majority of the specimens, and it appears to be very occasional in its occurrence.

The ventral beak in mature examples is small and strongly incurved. In one young specimen, in which the convexity of both valves is slight, and possibly in others, the beak is erect and there is a small area, probably with deltidial plates. A preparation exhibiting the interior, referred to this variety merely because it is more abundant than the nominal form, shows the presence of a long, slender, but low median septum in the dorsal valve and two dental plates in the ventral. The dental plates are thin. Their collocation is so far apart and their direction so divergent that they are at no point far from the apical walls, and in a specimen preserved in any other way might readily escape notice. The hinge plate is small and scarcely deserves the name, consisting of but little more than a thickening of the margin between the two dental sockets. The crura are short and strongly curved. They originate close together and rapidly diverge. At first they are directed nearly parallel to the plane of the valves, but later their curvature brings their direction nearly at right angles to it.

The structure of the area shown by the young specimen previously described is suggestive of Terebratuloidea, but the shell in that genus is plicated, the beak erect in mature forms not small and appressed, and the ventral valve is without dental plates. The internal structures are not far from those of Rhynchonella but this shell has no false area and has an incurved and appressed ventral beak.

The configuration of this and the related varieties is suggestive of Pugnax, and especially of those species which occur in the earliest Carboniferous time, no less in the general absence of plications than in the incurved beak. From the type represented by Pugnax osagensis they differ in these same characters. On the inside, however, they have a low septum and in this respect are much more like P. osagensis than the early Mississippian forms. The large indented hinge plate of Pugnax and Hypothyris is lacking. This form is not therefore in complete agreement with any of the three genera with which it appears most closely allied, and probably the institution of a new group would be justified did not the multiplicity of variation in exterior and structure shown by these forms demand a more extended investigation than I am able to undertake.

This variety and the species to which it is related are rather abundant in the "dark limestone," but only one specimen, which represents the variety, has so far come to hand from the Capitan formation. From this fact and from the lithology I am led to entertain some suspicion of the correctness of the locality label.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (stations 2965, 2969, 3500).

PUGNAX NITIDA n. sp.

Pl. XXIV, figs. 15 to 15c.

Shell small, transverse, subpentagonal. Ventral valve shallow over the posterior portion, strongly upturned in the anterior, which includes nothing but the sinus. The latter is broad and shallow. Its limits are not well marked over the posterior half. It is, however, defined along the margin by an angulation, the extension in front of which makes up all the anterior half of the shell. Beak small and erect (?).

Dorsal valve very transverse, nearly flat longitudinally, strongly arched transversely. Fold broad and high, not well defined, except along the margin, where there is a sharp upturning of the edge. The fold and sinus contain five or six plications, which are very indistinct at the margins and can not be seen at all farther back. There are no lateral plications, unless the indistinct grooves on the dorsal valve and the ridges on the ventral by which the fold and sinus are defined be considered as such.

The internal structures, so far as ascertained, consist of two dental plates in the ventral valve and a median septum in the dorsal. The generic identification, therefore, is not well established, but is probably limited to Pugnax and Camarotœchia. The configuration, however, is distinctly that of Pugnax.

The general appearance of the species is such as to lead one to imagine it the ancestor of the Pugnax? bisulcata group of the "dark limestone," but it is not yet safe to assert that they belong to the same genus. I have not seen in the literature any other species with which it is necessary to compare this one, though some of the figures of Camarophoria globulina given by Davidson resemble it.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2920).

PUGNAX SWALLOWIANA Shumard.

Pl. XV, figs. 8 to 12d; Pl. XXI, figs. 17 to 19?.

1859. Cameraphoria Swalloviana. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 394, pl. 11, figs. 1a to 1e (date of volume, 1860).

White Permian limestone; Guadalupe Mountains, Texas and New Mexico.

1897. Pugnax swallovana. Schuchert, Bull. U. S. Geol. Survey No. 87, p. 295.

Shell small, subpentagonal, flattened or quite gibbous, width equal to or greater than the length, widest near the front. Dorsal valve much more elevated than the ventral, but degree of elevation very variable; in some specimens a distinct depression in front of the beak, and sometimes one on either side forming two small lobes; mesial fold rising near the middle of the shell and becoming prominent in front, variable in width and height according to the number of ribs. Surface marked (in the specimen figured) with seven prominent subangular ribs, which commence about the middle of the valve and extend to the front; of these, three are situated on the mesial fold, and two on either side. Ventral valve gently rounded on the umbo, and marked on either side by a depression, which extends from the beak to the middle of the lateral edges; mesial sinus confined to the anterior half of the shell, broad in the middle and becoming narrow at the extremities, ornamented with two or more rounded plications, which are not as prominent as those limiting the sinus; beak elongated, sharply pointed, and slightly curved; shell structure distinctly fibrous.

This species is very variable in its characters. Some examples are extremely gibbous, and others much flattened. The ribs on the sides are quite distinct in some specimens and nearly obsolete in others. This species resembles in many respects C. globulina of Phillips, but it may be distinguished easily by its more elongated and sharply pointed beak and the less angular form of its ribs.

Dedicated to Prof. G. C. Swallow, State geologist of Missouri.

Formation and locality.—Permian rocks of the Guadalupe Mountains in Texas and New Mexico. It is one of the most abundant and characteristic species of the upper white limestone.

The foregoing is the original description of this species, which is one of the common forms in the Capitan limestone. Shumard calls attention to many of the variations, and they are quite as numerous as in its congener P. osagensis. The shape varies from subtriangular to pentagonal. There are usually three plications on the fold and two in the sinus, but in rare examples only two are found on the fold and one in the sinus. In a number of specimens the median plication is lower than those on each side of it. Of the lateral plications there are often three on each side, the final one being faint, especially on the dorsal valve, but sometimes only two appear to have been developed. The plications are usually high and angular, but are sometimes lower and rounded, the lateral ones being as a rule weaker than those on the fold and sinus. This tendency is carried so far in one specimen that though plainly marked by teeth on the line of junction the sides are entirely unplicated. Some of the specimens are explanate and have a triangular shape. Others are more contracted, globular, and with an acuminate-ovate form. Between these all gradations occur. As a rule the globular specimens have faint, rounded plications and the explanate high, sharp ones. I do not, however, at present regard these variations as deserving recognition.

The internal structures of this species consist of two short, stout dental plates in the ventral valve and a distinct though low septum in the dorsal. There is no spondylium, and Shumard's reference of the species to Camaroplloria is evidently a mistake. The external characters are strongly those of our common Pugnax osagensis, so much so that there is some doubt of the validity of Shumard's species. I retain P. swallowiana, however, because of its different faunal association, its larger size, and its plications, which are for the most part not so deep, while of those on the fold the central one tends to be smaller than the others. Furthermore, the septum is usually stronger. P. osagensis also frequently has a septum in the dorsal valve similar to the one in P. swallowiana, and this feature is perhaps more common in the genus, especially in Pennsylvanian species, than Hall and Clarke's diagnosis would lead one to infer.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

PUGNAX ELEGANS n. sp.

Pl. XV, figs. 13 to 14a.

Though originally including this form under Shumard's term Pugnax swallowiana, on revising the subject I find it desirable to discriminate it, and a separation of the specimens proves relatively easy. Shumard's figures fortunately leave but little doubt as to the variety to which the term swallowiana should attach, although his figures certainly are not very reliable.

The present form has a very spreading triangular shape, with a prominent fold and sinus. The plications in number and arrangement resemble those of P. swallowiana. There are three angular ones on the fold and two lateral ones on either side. It frequently happens that the median plication of the fold is distinctly smaller and lower than those adjacent to it. The lateral plications are rounder and fainter than those on the fold and are more distinct on the ventral valve than on the dorsal.

This species is distinguished from P. swallowiana by its more spreading, triangular shape, and by the more pronounced tendency of the median plication of the fold and sinus to be aborted.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906), Guadalupe Mountains, Texas.

PUGNAX SHUMARDIANA n. sp.

Pl. XV, figs. 15 to 17c.

Shell large, triangular, gibbous. Ventral valve rather flat though much turned upward toward the front. Sinus, as exhibited by a depression in the shell, broad, shallow, and somewhat well defined. Beak erect, acuminate. Sinus with four plications. There are three distinct lateral plications; the fourth is indicated chiefly by a denticulation at the margin.

Dorsal valve strongly convex. Fold, as an elevation of the surface, not very high, distinct toward the front, but merging into the general convexity about half way back. It is occupied by five plications. Lateral plications on each side three in number, with often a fourth indistinct one.

None of the plications extends to the beak, a feature rather characteristic of the genus. They are high and angular, those on the sinus being stronger and more persistent than the lateral ones. While the fold and sinus are not prominent as flexures of the shell they are very distinct and strong in the front view.

Internally this shell bears a low median septum and two rather stout, short dental plates.

I can not regard this as the same species with P. swallowiana. It is much larger, and has five instead of three plications on the fold.

The above description is based on four somewhat imperfect specimens from the Capitan limestone. A fifth example, probably a young specimen, is represented by figs. 17 to 17c of Pl. XV. It is instructive in many ways. The fold has five plications, the central of which is strongest and most extended. Those on each side are less prominent and the final ones are shortest and faintest of all. Those in the sinus correspond. Although the shell is rather large there are no lateral plications. The fold and sinus are practically undeveloped and to be detected only by a slight deflection of the contact of the valves in front. Both valves are very flat. This example well shows that the distinctive characters, except the plications of the fold and sinus, are of a late development, the convexity, lateral plications, and fold appearing as the shell nears maturity.

To this species I have referred with hesitation another somewhat fragmentary example from the same beds showing departures from the description given above. The plications are finer and more numerous. There are at least five on the fold, with five lateral ones on each side. It is possible that this may belong to a different species.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

PUGNAX OSAGENSIS Swallow.

Pl. XXIV, figs. 16 to 16b.

1852. Terebratula pugnus, Roemer Kreid (non Martin). von Texas, p. 89.

Carboniferous: San Saba Valley, Texas.

1858. Rhynconella (Camarophoria) Osagensis. Swallow, Trans. Acad. Sci. St. Louis, vol. 1, p. 219 (date of volume, 1860).

Upper "Coal Measures:" Missouri and Kansas.

1859. Rhynchonella uta, Meek and Hayden, Proc. Philadelphia Acad. Nat. Sci., p. 27.

Upper "Coal Measures:" Manhattan, Kans.

?1861. Rhynchonella sp. Salter, Quart. Jour. Geol. Soc. London, vol. 17. p. 64, pl. 4, fig. 5.

Carboniferous: Isthmus of Copacabana, Lake Titicaca.

1866. Camarophoria globulina. Geinitz, Carb. und Dyas in Nebraska, p. 38, pl. 3, fig. 5. (Not C. globulina Phillips.)

Upper "Coal Measures:" Bennetts Mill and Nebraska City, Nebr.

1872. Rhynchonella Osagensis. Meek, Final Rept. U: S. Geol. Survey Nebraska, p. 179, pl. 1, figs. 9a, 9b; pl. 6, figs. 2a, 2b.

Upper "Coal Measures;" Nebraska City, Nebr.

"Coal Measures:" Iowa, Missouri, Kansas.

Upper, middle, and lower "Coal Measures:" Illinois.

1873. Rhynchonella osagensis. Meek and Worthen, Rept. Geol. Survey Illinois, vol. 5, p. 571, pl. 26, fig. 22.

"Coal Measures:" Danville and Fulton County, Ill.

1884. Rhynchonella uta, White, Thirteenth Rept. Geol. Survey Indiana, p. 132, pl. 25, fig. 6.

"Coal Measures:" Indiana.

1891. Rhynchonella uta. Keyes, Proc. Philadelphia Acad. Nat. Sci., p. 247.

Lower "Coal Measures:" Des Moines, Iowa.

1893. Pugnax uta. Hall and Clarke, Nat. Hist. New York, Pal., vol. 8, pt. 2, p. 204. (Advance distribution in fascicles.)

"Coal Measures:" Manhattan, Kans.

1894. Pugnax uta. Hall and Clarke, Int. Study of Brach., pt. 2, pl. 44, figs. 17-19.

"Coal Measures:" Manhattan, Kans.

1895. Rhynchonella uta. Keyes, Rept. Missouri Geol. Survey, vol. 5, p. 103, pl. 41, fig. 7. (Date of imprint, 1895.)

Upper "Coal Measures:" Kansas City and Lexington, Mo.

1895. Pugnax uta. Hall and Clarke, Nat. Hist. New York, Pal., vol. 8, Pt. 2, p. 204, pl. 60, figs. 39-42.

"Coal Measures:" Manhattan, Kans.

1896. Rhynchonella uta. Smith, Stanford Univ. Publ., Cont. Biol. No. 9, p. 30.

Upper "Coal Measures:" Sebastian County, Ark., and Poteau Mountain, Ind. T.

1896. Rhynchonella uta. Smith, Proc. Am. Phil. Soc., vol. 35, p. 30.

Upper "Coal Measures:" Sebastian County, Ark., and Poteau Mountain, Ind. T.

1900. Pugnax Utah. Beede, Rept. Univ. Geol. Survey Kansas, vol. 6, p. 93, pl. 12, figs. 7-7c.

Upper "Coal Measures:" Bronson, Bourbon County, Kansas City, Iola, Olathe, Lawrence Lecompton, Topeka, Beaumont, Grand Summit, Kans.

1903. Pugnax Utah, Girty, Prof. Paper U. S. Geol. Survey No. 16, p. 412, pl. 7, figs. 14-14b.

Maroon formation: Crested Butte district, Colo.

This species occurs in considerable abundance in the black limestone at the base of the Guadalupe section, and appears to be identical with the common Pugnax osagensis of the Mississippi Valley Pennsylvanian. The shape is rather constant and is what may be described as triangular, with rounded basal corners. The fold and sinus are moderately strong and the plications are more or less blunted and rounded. There are regularly three on the fold and three on each side, the last lateral plication being rather faint. However, two specimens have been found with two plications on the fold, and one with four, though these are fragmentary and may not have been really conspecific with the others. The dorsal septum is faint and in some cases possibly absent.

I can not refer this species to P. swallowiana, from which it is without doubt varietally distinct. It has blunter plications, less variety in shape, and a fainter dorsal septum. In these characters it agrees With P. osagensis, though somewhat larger than the average of that species.

I have also referred here a single dorsal valve from the Delaware Mountain formation obtained at station 2931. It is slightly larger and more inflated than specimens from the black limestone. The plications are rounded. There are three on the fold and three on the sides, the final one being very faint. A distinct but probably not very high median septum occurs in this shell.

American paleontologists have for many years regarded Swallow's Pugnax osagensis as the same species as Marcou's Pugnax utah, and there is probably little doubt that they are closely related forms. Tschernyschew has recently, and I believe rightly, questioned the advisability of this course. If we accept Marcou's description and illustrations at their face value, and if we are prepared to make fine discriminations of species in this group, it must be admitted that the common Pennsylvanian Pugnax, to which the specific term osagensis was applied, is distinct from Marcou's figures of P. utah. Now, I think it highly improbable that Marcou's figures are to be relied on, but in view of the considerable difference of facies shown by the Utah fauna it seems to me inadvisable to continue to combine the two forms without some evidence that they are really the same. Consequently I am following Tschernyschew in reviving Swallow's species.

Horizon and locality.—"Dark limestone," Pine Spring (station 2930?); Delaware Mountain formation, Guadalupe Point (station 2931?); basal black limestone, Guadalupe Point (station 2920); Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2964).

PUGNAX BIDENTATA n. sp.

Pl. XXI, figs. 20 to 20c; Pl. XXIV, figs. 17 to 17c.

Shell small. Ventral valve rather strongly bent longitudinally and inflected at the sides. Sinus broad and deep, occupying most of the valve; defined by a rather sharp angulation, which forms, even where undefined, the line along which the shell is inflected. A slight groove lateral to the angulation adds to its prominence and is the only indication of lateral plication. The sinus contains a low median rib not visible except toward the front margin. Beak apparently rather large, high, and erect.

Dorsal valve strongly convex. Fold high, broad, occupying most of the shell; defined at the sides by an angular groove, which is succeeded laterally by a low ridge, making one indistinct lateral plication on each side. The fold bears a rather strong median sulcus, so that its summit is divided into two plications. The plications of both valves are subangular and moderately strong, but disappear a slight distance from the margin.

The typical example of this pretty species was found in the black limestone at the base of the Guadalupian section. Another very similar though somewhat larger example was found in the "dark limestone" at a much higher horizon. It shows some points of difference, as, for example, the presence of more distinct and more numerous lateral plications, but on the whole appears to belong to the same species as the other. There seems to be but little intergradation between P. bidentata of the black limestone and P. osagensis, with which it is associated; but of the mutation of P. swallowiana found in the "dark limestone" P. bidentata, occurring with it, might be considered as a mere variety. It would not surprise me, therefore, if further collecting would produce forms which are intermediate between P. osagensis, P. bidentata, and P. swallowiana. P. bidentata is closely similar to some of the forms figured as Camarophoria globulina. While I am by no means sure, I believe that the species under discussion is a Pugnax, in which case further comparisons with Camarophoria globulina will be unnecessary.

Horizon and locality.—"Dark limestone," Pine Spring (station 2930); basal black limestone, Guadalupe Point (station 2920), Guadalupe Mountains, Texas.

PUGNAX PINGUIS n. sp.

Pl. XXI, figs. 21 to 21c.

This species is proposed for a specimen from Shumard's "dark limestone" which, though similar to other species of the fauna, is yet too unconnected to permit its union with them. The shell is rather large, the length being 13 mm. and the width 14.5 mm. It has a subtriangular shape, with rounded basal angles. The ventral valve is rather shallow; beak small and erect; sinus broad, shallow, and distinctly defined, but not perceptible far back from the anterior margins.

Dorsal valve flattened over the central portions, strongly curved near the margins. Fold low, not distinct, except near the front. Plications rather thin and high, but rounded, dying out about halfway from the margins. There are four on the fold and three in the sinus, with four lateral ones on each side in the dorsal valve and five in the ventral, the final one in each case being faint.

This species is most closely related to Pugnax osagensis from station 2920 in the black limestone, but is larger and has more numerous plications.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

PUGNAX? PUSILLA n. sp.

Pl. XXIV, figs. 18 to 18b.

The following description is based on the single specimen which our collection contains. The small size of this specimen suggests that it is immature, but this inference is contradicted by its convexity and by the full development of its fold and sinus. The shape is broadly oval, the length somewhat greater than the breadth. The plications are narrow, high, and angular, extending completely to the beaks. There are three on the fold and four on each side. The fold and sinus are fairly well developed, especially in the anterior view.

The strength of the plications and their persistence to the beaks distinguish this form from most of those here referred to Pugnax. The nearest in this respect is P. pinguis, but the plications are much coarser, broader, and rounder in that species. In regard to the character of the plications, Rhynchonella longæva is perhaps nearer than any other Guadalupian species, but there are important differences in their number and arrangement.

Horizon and locality.—Basal black limestone, Guadalupe Point, Guadalupe Mountains, Texas (station 2967).

PUGNAX sp. a.

This species is represented by a single imperfect ventral valve whose characters do not permit it to be referred to any of the species recognized. In size and shape it is similar to Pugnax swallowiana, with which it is associated, but the shape is somewhat less triangular and the plications are finer and more numerous. They are rounded and die out before reaching the beak. There are two in the sinus and five on each side. The difference in size and arrangement of the plications does not permit this form to be placed with P. pinguis, unless intermediate stages not represented in our collection should prove to exist.

Horizon and locality.—"Dark limestone," Pine Spring, Guadalupe Mountains, Texas (station 2930).

Genus RHYNCHONELLA Fischer de Waldheim?

Internally the shells subsumed under this title do not differ materially from those which have been placed with Pugnax. They possess a pair of well-developed dental plates and a median dorsal septum, but if we overlook the single somewhat doubtful instance of an obscure cruralium and perforated hinge plate in the dorsal valve of one of the Rhynchonellas there are no practicable discriminating characters within it, so far as known. Externally a well-marked difference of configuration exists, the Rhynchonellas having numerous fine, persistent, somewhat bifurcating ribs, while in Pugnax the ribs are few, large, evanescent, and simple.

The Guadalupian Rhynchonellas are distinguished from the Carboniferous shells which foreign authors have assigned to Uncinulus chiefly by external characters, for though there are some differences in internal structure (such as the thickened ventral shell in Uncinulus, etc.), and perhaps indications of others, the main internal characters (the presence of a median septum in the dorsal valve and of dental plates) remain about the same in both. The plications in both genera are numerous and fine, but in Uncinulus they are more or less restricted to the margin and are indented by sulci, a character not known among the Guadalupian Rhynchonellas. The latter differ from Terebratuloidea in internal structure as well as configuration, since they have both dorsal septum and dental plates, possess finer and more numerous plications, and are without the round foramen and truncated beak of Waagen's genus. No comparison is necessary with Rhynchopora, since the Guadalupian Rhynchonellas possess a fibrous and not a punctate shell. Although there is considerable variance shown by authors in the employment of the generic term Rhynchonella, the Guadalupian shells seem to agree in most points with the species so identified.

In all, four Guadalupian species are included in this group, three of them already described by Shumard and the fourth apparently new. Two of the Shumard species have not been recognized in the more recent collections from the Guadalupe Mountains, and the four constituent types are at present so imperfectly known that I shall not attempt to assemble them into subordinate groups.

RHYNCHONELLA? INDENTATA Shumard.

Pl. XV, figs. 20 to 20c.

1859. Rhynchonella indentata. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 393 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

Shell variable, subovate, gibbous, length and breadth about equal, sides converging rapidly from the middle of the shell to the beak, and rounded toward the front, which is slightly indented. Dorsal valve strongly arched, much more elevated than the ventral valve; umbo flattened, broadly and rather deeply excavated in front by the tongue of the opposite valve; mesial ridge slightly elevated, and in some specimens scarcely perceptible except at the front; lateral margins very sinuous, being deeply indented at the cardinal margin on either side of the beak by the false area of the opposite valve; beak flattened and closely incurved. Ventral valve convex in the umbo and sides, scarcely gibbous, having a broad shallow sinus, which becomes obsolete on the umbo; false area well developed, distinctly defined, depressed below the plane of the dorsal valve, and marked with fine striæ; beak acute and moderately incurved. Surface marked with fine striæ of growth and from 20 to 25 rounded radiating costæ, which become obsolete on the umbo; costæ in the mesial fold and in the sinus.

Dimensions.—Length, 0.55; width, 0.50; thickness, 0.40.

Locality—White limestone of the Guadalupe Mountains.

Of this interesting species I have but four specimens. Two of these are from the "white limestone" and two from the "dark limestone." The latter are rather small and fragmentary. One of the other examples is also fragmentary, but the third is, fortunately, very perfect. There can be no doubt that the two examples from the white limestone belong to Shumard's species, and they permit me to add somewhat to his in the main faithful description. The figured specimen has about 25 plications, nine of which are on the fold and the rest lateral. There are five plications on the top of the fold and two on each side of it. Nevertheless this feature can be said to be moderately high, quadrate, and well defined. The surface is crossed at regular intervals by sublamellose concentric lines, of which no mention is made in Shumard's description. The beak is erect and appears to have a rounded foramen, closed below by deltidial plates. The flattening of the umbo of the dorsal valve is a striking character and appears at first to be artificial, but is repeated in all the specimens. The false area of the ventral valve is also important. It is defined by a distinct angular line. The dorsal valve has a somewhat corresponding flattened lateral area, but it is narrower and not distinctly marked off from the rest of the shell.

The internal characters have not been entirely made out. It has been ascertained, however, that the ventral valve possesses two slender, rather short dental plates, and the dorsal a long, slender, but not very high median septum. It is evident that the internal structures of this species prove Shumard's original assignment to Rhynchonella to be very nearly correct. Though the structure of the ventral beak suggests Terebratuloidea, that genus is without dental plates, and the general expression of the plications, surface ornamentation, false area, and erect beak are rather those of the true Rhynchonellas. Though the type of Rhiynchonella, according to Hall and Clarke, is without a septum in the dorsal valve. Waagen names this as one of the characters in his definition of that genus, and it is only in the sense that Waagen used the name that this species can be referred to it.

The concentric striation, erect beak, flattened umbo, and false area, taken in connection with the internal structures, give this species marked individuality, and appear not to exist in combination in any of the described genera. It is possible that R. indentata is representative of a new group. This, however, will not admit of satisfactory determination until the internal characters are better known.

The smaller specimens from the "dark limestone" have a less distinct false area, and one of them shows but four plications on the top and but one on the sides of the fold. There are about eight lateral plications on either side.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

RHYNCHONELLA? LONGÆVA n. sp.

Pl. XV, figs. 18 to 18b, 19 to 19c.

My material representing this species is rather fragmentary, though it is without much doubt distinct from anything so far known from the Guadalupe Mountains. The following characters have been ascertained:

The shell is small and subtriangular. The lateral and front margins are nearly straight; the anterolateral angles broadly rounded. The ventral valve is comparatively shallow, turning up strongly in front in a broad, ill-defined sinus. The beak is probably erect and pointed. The dorsal valve is rather gibbous. Fold not well defined, except at the front. Plications 17 to 20 in number, rather angular, high, and reaching to the beaks, increasing by bifurcation. About four plications occur on the top of the fold and one on each of its sides. There are six or seven lateral plications on either side.

In the interior of the ventral valve there are two dental plates not united into a spondylium but discrete, as in Pugnax. The general expression, however, is very different from that of Pugnax, and the exact position which this form should assume among the Rhynchonellidæ is uncertain.

In general appearance this species is not unlike Rhynchonella indentata, though it is conspicuously smaller and lacks the false area and flattened dorsal umbo which are such marked characters of that type. Similarly, it can not be referred to other species which I have placed with Rhynchonella? Compared with Camarophoria venusta it is much smaller, somewhat differently shaped, and has deeper and stronger plications.

The foregoing description is based on a few imperfect examples from the Capitan limestone obtained on El Capitan Peak. To the same species has been somewhat provisionally referred a single, rather crushed, example from about the same horizon on the foothills southwest of Guadalupe Peak (station 2906). Still more doubtful is the identification of two small specimens from the black limestone at the base of the Guadalupe section. They seem to have somewhat coarser and less numerous plications, and I suspect may be representatives of the genus Pugnax, though distinct from any of the species so far discriminated.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926); base of Capitan formation, hill southwest of Guadalupe Point (station 2906?); basal black limestone, Guadalupe Point (station 2920?), Guadalupe Mountains, Texas.

RHYNCHONELLA? GUADALUPA Shumard.

Pl. XVI, figs. 10 to 10b.

1858. Rhynchonella Guadalupae. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 295 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

1859. Rhynchonella Guadalupae. Shumard, idem, p. 392, pl. 11, figs. 6a to 6c.

White [Permian] limestone: Guadalupe Mountains.

Shell subtriangular, with the angles rounded, convex, wider than long; lateral margins nearly straight, converging at an angle of about 85°; sides presenting a large, well-defined, elliptical, concave, or flat, smooth area, which is carinated at the commissure of the valves and extends from the beaks nearly to the front; front strongly or slightly sinuate. Ventral (receiving) valve not as prominent as the opposite one; umbonal region flattened convex, having a broad, shallow mesial sinus extending from beak to front, lateral edges gently arcuate; beak flattened convex, rather strongly incurved. Dorsal valve presents a regularly convex and rather gentle curve from beak to front and a low, broad mesial elevation, which is scarcely perceptible except near the front; beak depressed, gently convex and closely incurved. Surface marked with numerous, rather coarse, rounded, radiating striæ, their number increased by bifurcation and insertion. The bifurcations generally take place near the beak. At the border the number of striæ amount to from 30 to 35 on each valve.

Dimensions.—Length, 0.58; width, 0.76; thickness, 0.48.

A handsome species and quite characteristic of the white limestone of the Guadalupe Mountains of New Mexico and Texas.

This form is evidently related to Rhynchonella? indentata and R.? texana, and, like the latter, does not occur in our collections from the Guadalupe Mountains. Fortunately, Shumard has left a number of figures which ought to supplement his description and aid in the identification of it. The figures are reproduced on Pl. XVI of the present work.

Although no authentic material has come to hand which could be referred to this species, a single specimen from the Glass Mountains has been somewhat hesitatingly placed here. This specimen I have been so unfortunate as to lose, but as it presented some structural features of considerable interest I shall venture to give a brief account of it from memory. The specimen was silicified and represented only the posterior portion of a dorsal valve. Externally the shell was covered with a large number of very fine, flat, radiating ribs. Their general character recalled the sculpture of Orthotetes guadalupensis and differed so much from that of either Rhynchonella? indentata or R.? longæva that I felt unwilling to identify it with them. On the other hand, it was very suggestive of Shumard's figures of R.? guadalupæ. On the interior the apical portion was closed by a rather small horizontal hinge plate. From the bottom of the valve rose a thin, moderately high median septum, whose posterior portion subdivided above to form a very small cruralium. This little channel along the upper edge of the septum would naturally be blocked by the hinge plate, through which, however, it was continued by a small circular perforation. This perforation of the hinge plate is a character which has, I believe, seldom been observed in any group of brachiopods and should not, I thought, fail to be noted here, in spite of the imperfect and unfortunate character of my observations.

Horizon and locality.—Top of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2966?).

RHYNCHONELLA? TEXANA Shumard.

1859. Rhynchonella Texana. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 393 (date of volume, 1860.)

Dark [Permian] limestone: Guadalupe Mountains; conglomerate at the mouth of Delaware Creek, Texas.

Shell small, ovate, moderately gibbous, front rounded or slightly indented; anterolateral margins rounded, converging posteriorly to the beak at an angle of 63°. Dorsal valve as broad as long, convex, more elevated than the ventral valve, smooth on the umbo and having a distinct mesial elevation in front, bearing usually three prominent subangular ribs, on either side of which are two or three less prominent ribs; cardinal edges rather deeply indented by the false area of the opposite valve. Ventral valve gently convex, most prominent about the middle, marked in front with a moderately deep mesial sinus, which becomes obsolete on the umbo and bears two or more subangular ribs; beak prolonged, acute gently; foramen narrow, triangular; false area rather strongly developed, not very distinctly defined.

Dimensions.—Length, 0.35; width, 0.30; thickness, 0.22.

Formation and locality.—From the dark limestone forming the base of the Guadalupe Mountains, and from the conglomerate at the mouth of Delaware Creek.

This description represents a form in some ways so similar to Pugnax osagensis that it might be thought the shells from the "dark limestone" which I referred to that species really belong to this. Shumard, however, associates R. texana with R. indentata in the same genus, and some parts of his description seem to justify this association. It is evident, at least, that R.? texana had a false area similar to R. indentata, a character which is quite foreign to the group of Pugnax. I especially regret that my collections fail to contain this species, since Shumard neglected to figure it, and except for his description it is practically unknown.

Family TEREBRATULIDÆ Gray.

The Guadalupian Terebratulidæ have been assembled into four generic groups. One of these it has seemed necessary to describe as new, under the title Heterelasma. Another has been referred with some reservation to Notothyris, while the other two have been assigned to Dielasma and Dielasmina. Among the Dielasmas two groups can be distinguished, one of which is represented by Dielasma? scutulatum, the reference of which to Dielasma, or even to the Terebratulidæ, is open to more or less question. The other group includes four species, and might itself be subdivided, did such a course seem desirable.

If this representation is compared with that of the Salt Range of India very considerable differences appear. Waagen recognizes Dielasma, Dielasmina, Hemiptychina, and Notothyris, Hemiptychina being absent from the Guadalupian fauna and Heterelasma from that of the Salt Range. Including doubtful species, I have found only seven varieties of Dielasma in the Guadalupian, whereas Waagen recognizes ten in the Salt Range fauna. Of these, D. guttula, representing the group of D. sacculus, and D. elongatum and D. nummulus, representing the group of D. gillingense, seem to be unrepresented in the Guadalupian. The group of D. ficus, with three species, is probably represented by D. spatulatum and D. prolongatum. Waagen's D. truncatum seems to be somewhat intermediate between the two Guadalupian species mentioned, while D. itaitubense is rather different from either. The group of D. biplex with its four species, has somewhat the configuration of the species belonging to Heterelasma, but there are no Guadalupian Dielasmas with which it can be compared. On the other hand, D. sulcatum and D. cordatum, together with the doubtful form D.? scutulatum, have no closely related Indian species.

Dielasmina guadalupensis resembles to a considerable extent D. plicata of the Salt Range, but the American species is less abundantly plicated, and should probably be considered as racially somewhat immature.

In contrast to the Guadalupian, which contains only three varieties of Notothyris, Waagen found no less than eight species in the Salt Range. Notothyris simplex and N. inflata are most nearly related to the Guadalupian N. schuchertensis. The other Salt Range species are more or less widely different, and are perhaps more highly developed representatives of the genus. There is nothing in the Indian fauna which resembles Notothyris sp. found in the Glass Mountains.

In his first report on the Chitichun fauna No. 1 Diener recognizes one species of Dielasma, three of Hemiptychina, and two of Notothyris. The Dielasma is more or less related to D. spatulatum, but the single species of Notothyris of which figures are given (N. triplicata) differs considerably from the Guadalupian representatives of the genus.

In a subsequent paper on this fauna Diener distinguished one species of Hemiptychina, three of Dielasma, and four of Notothyris. His figure of Notothyris triplicata,a prepared to show internal structures, clearly represents what I think must be two dental plates, the presence of which would debar this species from Notothyris without doubt. The two species N. mediterranea and N. exilis appear to be closely related to N. schuchertensis, but the singular form described as N. walkeri is unlike not only the Guadalupian species but also the typical form of the genus. The configuration somewhat suggests Cryptacanthia. The three species of Dielasma do not seem closely allied to the congeneric Guadalupian forms. They are nearest to D. spatulatum and D. prolongatum. The form called Dielasma sp. indet. aff. hastæforme and D. elongatum suggest, in configuration at least, the doubtfully placed D. scutulatum.

---

aMem. Geol. Survey India, Pal. Indica, ser. 15, vol. 1, pt. 5, 1903, pl. 2, fig. 12.

---

The paper by this author on the Anthracolithic fossils from Kashmir and Spiti contains references to but one species of the Terebratulidæ, which is identified as Dielasma hastatum. It is probably somewhat related to D. cordatum, but not closely. When treating this fauna subsequently Diener recognized no terebratuloids in the lower fauna and only Dielasma latouchei in the upper. The Permian fauna from Kumaon and Gurhwal contains but an undetermined species of Dielasma, which somewhat suggests D. spatulatum.

From Malla Sangcha, Diener cites one species of Hemiptychina and five of Notothyris. N. mediterranea is the most closely allied to N. schuchertensis of these forms, and N. triplicata and N. minuta the least. There seems to be nothing comparable to Notothyris sp.

Dielasma latouchei, the only terebratuloid from the Lissar Valley, is quite unlike any Guadalupian representatives of the genus. The same species is recorded from Byans.

Two species were cited by Davidson from Kashmir as Terebratula sacculus Martin and T. austeniana n. sp. Neither seems to be comparable to any of the Guadalupian terebratuloids.

Considered as a whole, the terebratuloids of the Salt Range and Himalaya seem to me to show no very close relationship to the Guadalupian ones. The genera represented are in part different, and where the genera are the same many of the species belong to different groups, yet to say that they are not related at all will be going too far.

The only terebratuloid found by Kayser among the Lo Ping collections was referred to Terebratula hastata. It seems to have no closely allied Guadalupian form. From Kantschoufu, Loczy cites Dielasma vesiculare, represented by a little shell which does not seem closely related to any of the Guadalupian terebratuloids.

Rothpletz cites a species of Hemiptychina in his paper on the Permian of Timor and Rotti, calling it Terebratula himalayensis var. sparsiplicati. His figures show a form having the general appearance of Notothyris schuchertensis, but apparently the two species are not generically related.

From this it would appear that the Terebratulidæ are surprisingly scarce among the faunas of eastern Asia, and display no very marked resemblance to their Guadalupian congeners.

The Terebratulidæ of the "Permo-Carboniferous" of Queensland and New Guinea comprise, according to Etheridge, only four species of Dielasma. They are of the usual simple type, and though one or two of the Guadalupian Dielasmas resemble them, only a very remote relationship can be inferred from this group of Brachiopoda.

The Carboniferous Terebratulidæ cited from New South Wales by De Koninck comprise only two varieties, to which must be added the form described as Rhynchonella inversa. One of these forms was found only in the lower portion of the series; another, Terebratula hastata, which resembles in a general way Dielasma prolongatum, appears to have been found in both series. The third form, described as a Rhynchonella, also belongs to the earlier fauna.

From the Moskovian of Russia Trautschold records only one species, which he identifies as Terebratula sacculus. If we assume that it is a Dielasma, it has somewhat the specific characters of D. prolongatum without being very close either to that species or to D. spatulatum.

The terebratuloids of the Gschelian fauna are subdivided by Tschernyschew into the genera Dielasma, Hemiptychina, Notothyris, Aulacothyris, and Waldheimia. Of these, Hemiptychina, Aulacothyris, and Waldheimia are not known from the Guadalupian, while Heterelasma and Dielasmina of the latter appear to be absent from the Russian horizon.

No less than 14 species, according to Tschernyschew, occur among the Gschelian Dielasmas. The group of D. sacculus, including D. supracarbonicum and D. maelleri, shows distinct relationship to D. prolongatum. The group of D. gillingense, to which Tschernyschew refers seven species, also resembles D. prolongatum, though less closely, while some of the Russian forms can be compared to D. spatulatum. One species (D. plicatum) even resembles D. sulcatum. The group of D. ficus contains two species. D. truncatum shows some resemblance to D. prolongatum and D. itaitubense to D. scutulatum, but in neither case is it very close. Lastly, the group of D. biplex, comprising three species, seems to be unrepresented in the Guadalupian. All the Guadalupian species, with the exception of D. cordatum, seem to find more or less related forms in the Gschelian, only one group of three species being strictly peculiar to the Russian fauna. Among the species of the group of D. gillingense Tschernyschew recognizes the two American types D. millipunctatum Hall and D. bovidens Martin. I agree with Tschernyschew in discriminating these two species, which are united by most American paleontologists, for Hall's figures certainly show important differences from typical D. bovidens, and it is probable that the differences are real ones; but I must confess that, to judge by his figures, the Russian form which Tschernyschew calls D. bovidens does not seem to me by any means specifically identical with the Pennsylvanian species.

Tschernyschew recognizes four species of Hemiptychia in the Gschelian, only one of which, that identified as Hemiptytchina aff. pygmæa, appears to me to have the expression of veritable Hemiptychina. This type seems to be wanting in the Guadalupian, and I need not dwell on the group save to notice how much some of Tschernyschew's figures of H. orientalis recall the Guadalupian species Dielasma sulcatum in certain respects. D. sulcatum, furthermore, can not be certainly assigned to Dielasma.

Three species represent the genus Notothyris in the Gschelian fauna. N. nucleolus especially resembles the Guadalupian N. schuchertensis, the two other species less so. Some of the figures of N. nucleolus, however, somewhat suggest Notothyris sp., which is probably widely different from N. schuchertensis.

To Aulacothyris Tschernyschew assigns two Gschelian species. This author suggests that Aulacothyris Douville and Cryptacanthia are one and the same. If his references of the Gschelian species to Aulacothyris is correct, this opinion is probably well founded, for A. uralica in its peculiar configuration is strikingly suggestive of Cryptacanthia compacta. It is hard to believe that the resemblance is merely one of parallel development. This type is entirely lacking to the Guadalupian.

The species identified as Waldheimia pentagonalis is also quite unlike any known Guadalupian form.

The other papers relating to the Gschelian fauna which I have examined add so little to what Tschernyschew has given in regard to the Terebratulidæ that I shall pass them by and proceed to the consideration of the Artinskian terebratuloids.

Tschernyschew cites from the southern Urals only Dielasma prolongatum, a species which the Guadalupian D. spatulatum and D. prolongatum somewhat resemble. Stuckenberg cites D. plica, D. elongatum, D. uralicum, D. sacculus, D. seminula, and Dielasma sp., all of them, with one exception, unfigured. I will not, therefore, make individual comparisons with Guadalupian species, merely calling attention to the fact that we seem to have here only two genera represented, Dielasma and Aulacothyris (as Dielasma uralicum), the latter of which is not found in the Guadalupian, which has, on the other hand, Heterelasma, Notothyris, Dielasmina, and certain types of Dielasma which do not occur in the Artinskian. From the Kungurstufe Stuckenberg cites only Dielasma plica and D. elongatum. Krotow records from the Artinsk almost exactly the same series of species as Stuckenberg, viz, Terebratula plica, T. hastata, T. elongata, T. sacculus, T. vesicularis, and T. uralica, and the same remarks apply to them. Krotow's identifications also are unfigured. Sibirzew mentions as occurring in the Artinsk D. vesiculare and D. elongatum.

From the Permian of Russia, Tschernyschew records only Dielasma elongatum and D. sacculus, two species whose nearest allies in the Guadalupian are probably D. prolongatum and D. spatulatum. Netschajew from this horizon cites D. elongatum and D. angustum, of which the latter seems to be without any closely related Guadalupian species. Sibirzew mentions D. elongatum, D. sufflatum, and Dielasma cf. sacculus from the lower beds of the Permian and D. elongatum from the upper. Golowkinsky cites only Terebratula elongata.

To judge by the records which I have seen, the Terebratulidæ of the Russian section are reduced from very numerous species belonging to a number of genera in the Gschelian to a relatively few species, representing but two genera, in the Artinsk, while in the Permian only the genus Dielasma remains, represented by two or three simple, poorly characterized types. In the ample differentiation of its terebratuloids the Guadalupian fauna certainly possesses most points of comparison with the Gschelian, though many of the generic and specific types are different. The Permian terebratuloids resemble those of the Guadalupian only in the presence of a few simple forms of Dielasma, most of the Guadalupian genera and species being unrepresented in the Russian Permian.

From Djoulfa, in Armenia, Abich cites only Notothyris djoulfensis, a species quite distinct from the Guadalupian representative of the genus, though not unrelated to it.

Gemmellaro divides his Sicilian terebratuloids into the genera Rhaetina, Hemiptychina, and Rostranteris. Curiously enough, the almost universally distributed genus Dielasma seems to be absent from this fauna. Only a single species is referred by Gemmellaro to Rhaetina, and though it presents striking analogies in configuration to the Guadalupian species Heterelasma shumardianum, I am forced to believe that they possess no intrinsic relationship, since Heterelasma shumardianum has the structures of neither valve as defined by Waagen for Rhaetina.

Five Sicilian species of Hemiptychina are discriminated. Unfortunately, this genus is at present unknown in the Guadalupian fauna. Rostranteris, which Gemmellaro describes as a new genus, seems now to be generally regarded, by those who are in a position best to know, as a synonym of Notothyris. Gemmellaro recognizes ten species. Several resemble Notothyris schuchertensis and N. schuchertensis var. ovata, especially Rostranteris inflatum, but many of them belong to distinctly different groups. Two or three (e. g., Rostranteris ovale) to a certain extent recall the imperfectly known Notothyris sp. of the Guadalupian.

Somewhat in contradiction to what appears to be the case in other groups, the Terebratulidæ of the Guadalupian have but little in common with those of the fauna from Palermo.

In his paper on the fauna of the Carnic Fusulina limestone Schellwien recognizes only two terebratuloids, which he designates as Dielasma? carinthiacum and Dielasma? toulai. The former is a simple type, somewhat resembling D. prolongatum, but the latter appears to be unlike any of the Guadalupian species. The family seems to be much better represented in the Trogkofelschichten, where Schellwien recognizes eight species and four genera. To Notothyris he refers two species, both of which more or less resemble N. schuchertensis, especially N. ovalis. Dielasma is represented by only a single unidentified species of the general type of D. prolongatum, and in this fauna, as in the closely related Sicilian one, the scarcity of Dielasma is a somewhat noteworthy feature. Hemiptychina is accorded four species. This genus is not at present known in the Guadalupian, but H. carniolica of the Alpine fauna in configuration strongly suggests the form which I have described as Dielasmina guadalupensis. The last generic type from the Trogkofel is identified with White and St. John's genus and species Cryptacanthia campacta. The identification is queried; but the resemblance, it must be confessed, is close. Nothing of the sort is known in the Guadalupian. In general the Guadalupian Terebratulidæ seem to show no very close resemblance to those of the Trogkofelschichten. Gortani cites from the Carnic Alps only Notothyris epilis and Dielasma elongatum, both of them unfigured.

In his monograph on the German Dyas, Geinitz cites only Dielasma elongatum, the paucity of terebratuloids in this fauna allying it with the typical Permian. In the description of plates, the series of forms included under this general title are subdivided as D. elongatum, D. elongatum var. sufflatum, and D. elongatum var. latum and complanatum. D. elongatum itself resembles D. prolongatum and D. spatulatum of the Guadalupian, but the varieties latum and complanatum have no closely related species in that fauna, being perhaps nearest to D. spatulatum.

King identifies his terebratuloids from the a Permian of England as Epithyris elongata and E. sufflata. Both seem to be simple forms of Dielasma, related, but only in a general way, to D. prolongatum and D. spatulatum.

In the reduction of the terebratuloids to the single genus Dielasma, and the restriction of that genus to primitive and slightly differentiated forms, the Permian of England, the Dyas of Germany, and the typical Russian Permian are in agreement with one another, but in sharp contrast to the Guadalupian.

Among his fossils from the south point of Spitzbergen, Toula mentions a species of terebratuloid which he identifies as Terebratula hastata. It may be compared especially to Dielasma prolongatum among Guadalupian forms, but is probably not closely related. Lundgren also records an unidentified species from Spitzbergen, which he cites merely as Terebratula? sp. but does not figure. No other notices of the occurrence of this group in Arctic works have come under my observation.

Under the title Terebratula cf. gillingensis Stache cites a species from a fauna in the West Sahara which is probably considerably older than, as it is considerably unlike, the Guadalupian. The species as figured is only remotely related to any Guadalupian form.

From Bolivia, Toula cites Terebratula hochstetter, his figures representing a form more or less resembling Dielasma spatulatum and D. prolongatum. In Terebratula titicacensis, which Gabb described from Peru, I do not see a close relationship with any Guadalupian species. The general outline is something like Dielasma? scutulatum, but it is doubtful if the two species are related even generically.

Derby records from Brazil two species of terebratuloids which he cites as Terebratula itaitubensis and Waldheimia coutinhoana. The former has already been recognized as a Dielasma, but is not like any of the Guadalupian representatives of the genus. The Waldheimia also appears to be non-Guadalupian.

If we disregard a few species which are quite unknown save for the first description, unaccompanied by figures, the terebratuloids of the typical Pennsylvanian consist of only two species of Dielasma and one of Cryptacanthia. The Pennsylvanian species D. bovidens is comparable to D. spatulatum of the Guadalupian, though quite distinct specifically. D. obovatum, on the other hand, seems to have no allied Guadalupian species. The genus Cryptacanthia is likewise, so far as known, alien to the Guadalupian fauna. The Guadalupian terebratuloids contrast with those of the Pennsylvanian not only in the more varied generic differentiation, but also in the greater differentiation of the single genus which they possess in common, and in the occurrence of different species in each.

Genus DIELASMA King.

The generic position of the species referred to this genus has in no instance been completely demonstrated. For the most part the general expression and the presence of dental plates render the reference to Dielasma at least a very probable one, but in the case of Deilasma? scutulatum the entire uncertainty as to internal structure and the additional doubt as to the punctate character of the shell render this species an extremely problematical member of the group.

Four species are with considerable confidence placed with Dielasma, and they might perhaps with advantage be subdivided into as many groups, as they show considerable difference in character.

DIELASMA SPATULATUM n. sp.

Pl. XVI, figs. 3 to 4c.

?1859. Terebratula elongata a Shumard (non Schlotheim); Trans. Acad. Sci. St. Louis, vol. 1, p. 392 (date of volume, 1860).

[Permian]: Guadalupe Mountains.

---

aA specimen from shumard's collection identified, probably by him, as Terebratula elongata var. sufflata is the dorsalvalve of Squamularia guadalapensis.

---

Shell rather large and spatulate. Greatest width near the anterior margin. Ventral valve shallow in its transverse curvature, more strongly bent longitudinally. About midway the shell, which has been convex in its posterior region, becomes slightly concave, producing a broad, shallow, undefined sinus, which toward the front occupies most of the width. Beak probably large and projecting.

Dorsal valve nearly flat longitudinally, gently and evenly arched transversely. There is no fold aside from the regular curvature of the shell, a circumstance which, conjoined with the broad sinus of the opposite valve, produces a truncation of the front margin.

On the interior the dorsal valve shows a median septum, with which are probably associated the other plates of Dielasma. The ventral valve has two strong dental lamellæ.

This species resembles D. truncatum Waagen, but can hardly be the same thing.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966?); Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Point (station 3762d?), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2962?).

DIELASMA PROLONGATUM n. sp.

Pl. XVI, figs. 5 to 5c.

Shell rather large, elongate-ovate. Greatest width about one-third the length back from the front margin. Ventral valve strongly bent longitudinally. Transverse curvature convex over the posterior third, gently concave farther forward. Toward the front the depression deepens and strengthens into a strong but ill-defined sinus. Both valves, but the ventral one especially, have a flattened reflex rim along both sides.

Dorsal valve nearly flat longitudinally, rather strongly convex transversely, with a flattened rim or margin at the side. No fold distinct from the general curvature. Owing to the configuration of the two valves, the front end is somewhat emarginate. The posterior end is long and pointed, with the beak hardly perceptible as a distinct feature.

This form at first seems to be very distinct from D. spatulatum by reason of its flattened margins, more elongate shape, and deeper sinus; but the former character rather suggests that the individual may represent a senile condition. The growth lines of the specimen are unfortunately not visible to serve as evidence, but it seems probable that at a somewhat earlier period the specimen might not have differed greatly from D. spatulatum. Should this prove to be the case, the present form would probably be no more than a variety of that species. In its present development, however, it appears to be more nearly related to D. sulcatum, but should be readily discriminated by reason of its larger size, more regularly ovate shape, and broader and rounder sinus.

Horizon and locality.—Capitan formation, McKitterick Canyon, Guadalupe Mountains, Texas (station 2932). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

DIELASMA CORDATUM n. sp.

Pl. XVI, figs. 2 to 2c.

The typical specimen of this species has the dorsal valve much depressed or flattened, especially over the anterior half of the shell, so that the space between the two valves is unusually slight. In the ventral valve the sinus, which extends less than half the shell length back from the front margin is, though not perceptibly deeper than in D. spatulatum, much narrower, producing a distinct indentation in the anterior outline. Otherwise, except where the expression is altered by the variations mentioned, it is much like that of D. spatulatum. The narrow sinus and the emargination which it produces are very suggestive of D. prolongatum, and in a less degree of D. sulcatum. These four forms are evidently related, but my collections are not complete enough to indicate how close. I can by no means at present regard any two of these forms as specifically the same.

Dielasma cordatum simulates some varieties of Composita emarginata very closely, but of course the resemblance is extremely superficial.

Horizon and locality.—Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

DIELASMA SULCATUM n. sp.

Pl. XVI, figs. 1 to 1c.

Shell of medium size, subpentagonal. Ventral valve moderately convex. Beak as usual in the genus. Greatest width about two-fifths the length back from the anterior margin. The sinus can be traced nearly to the beak as a depressed line, but near the front it rapidly expands, becoming deep and subangular.

Dorsal valve moderately convex. There is no distinct fold, but the valve as a whole has the shape of a dihedral angle, the sides of which are more nearly plane toward the front, producing a distinctly peaked shape. Farther back the shell is more arched and the angulation is lost. The ventral sinus is so strong that a well-marked emargination is produced at the front margin.

The interior of the species is not known and its generic position is therefore uncertain. The punctate shell and configuration demonstrate its relation to the terebratuloids and it can probably be placed with Dielasma.

Horizon and locality.—Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

DIELASMA? SCUTULATUM n. sp.

Pl. XVI, figs. 8 to 9b.

Shell small, elongate. Ventral valve shallow, most strongly convex in the posterior half. Beak small, produced, rather strongly incurved. Shape rhombic, greatest width about midway. Posterior end pointed and tapering; anterior contracted and strongly rounded.

Dorsal valve shallow but more convex than the ventral. The curvature across the shell is less than that lengthwise. The dorsal valve has not an appreciable fold nor the ventral a perceptible sinus, but a flexure of their plane of union at the anterior margin indicates that these features are present if not obvious.

The surface appears to be smooth and the shell substance fibrous, doubtfully punctate.

I hardly know where to place this little shell generically. Only three specimens, one of them somewhat imperfect, have come to hand and it has not been possible to ascertain any of the internal structures. The general expression, and especially the configuration of the ventral valve, strongly suggest a relationship with the terebratuloids. The shell substance, however, is distinctly fibrous, instead of foliaceous, and it is uncertain whether it is punctate or solid. Certain local areas have the appearance of being obscurely punctate, but I am not altogether sure that this appearance is not connected with the prismatic fibrous structure of the shell substance. At one time I doubtfully placed these forms with Spirigerella, but the beak in none of the specimens is sufficiently perfect to show plainly whether it is truncated by a large foramen or incurved and pointed. The preservation in the Guadalupe Mountains is such as sometimes to obscure the invertebrate shell structure, and it has seemed less liable to be erroneous to place this species with the terebratuloids, though its position is still subject to revision. The general shape of this species is that of a young and elongate specimen of Maartinia rhomboidalis, but the form of the beak seems to preclude any possibility of its being a Martinia.

Horizon and locality.—Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

Genus DIELASMINA Waagen.

This genus differs from Dielasma, so far as known, only in configuration, for instead of having a well-marked fold and sinus on the dorsal and ventral valves, sometimes modified by the development of a subordinate mesial sulcus and plication, Dielasmina possesses a series of nearly equal submarginal plications without any distinct fold or sinus. The shape of the one type is apt to be flattened and of the other globose.

Dielasmina guadalupensis seems to meet these conditions in every respect. While it differs from the genotype, D. plicata, in having the plications larger, fewer, and more marginal, it appears to be much more closely allied to D. plicata than to the usual type of Dielasma, and I have placed it with some confidence in Waagen's genus. The relations of Dielasmina? perinflata are much more uncertain, for Shumard's species has never been figured and its general character is imperfectly known.

DIELASMINA GUADALUPENSIS n. sp.

Pl. XVI, figs. 6 to 7a; Pl. XXI, figs. 22 and 22a.

Shell of medium size, inflated, ovate. Ventral valve strongly convex. Beak large, much incurved. In mature specimens there is a well-marked median sulcus, on each side of which is a smaller sulcus separated by a subangular plication, beyond which, on the type specimen, still another low plication is seen. The plications and sulci are practically marginal and can be traced but a short distance back from the edge of the shell.

The dorsal valve is moderately convex. Its plications correspond to those of the ventral valve. It has a strong mesial plication bonded by two subangular sulci, beyond which on each side is a less distinct plication followed by a very indistinct sulcus.

Of this species I have seven specimens—none of them very perfect—two from the Capitan limestone and five from the "dark limestone." In the plications, which constitute the most striking feature of this species, considerable variation is exhibited, depending largely on difference in size. A large ventral valve from the Capitan limestone shows only a shallow median sinus, its two bounding folds, and a shallow depression on either side, while in a smaller individual only the sinus is developed. In this condition it is practically impossible to distinguish specimens from young Dielasma sulcatum. In a general way these shells are very suggestive of some of Waagen's figures of Dielasma biplex and Dielasma problematicum; but, unfortunately for comparison, the ventral valve of my specimens has a median sulcus, instead of a median plication. It is more loosely and less abundantly plicated than the Indian species of Dielasmina. I am, however, in some doubt as to whether this may not really be Shumard's D. perinflata. It is evident from Shumard's description that his form did not have the plication developed to anywhere near the degree shown in D. guadalupensis, but this feature varies considerably in D. guadalupensis, so that D. perinflata may be based on small or imperfectly developed specimens of my species. It seemed to me more prudent, however, to employ for the present group a distinct name, leaving it to be settled from further collections in what relation it stands to D. perinflata.

Horizon and locality.—Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930), Guadalupe Mountains, Texas.

DIELASMINA? PERINFLATA Shumard.

1859. Terebratula perinflata. Shumard, Trans. Acad. Sci. St. Louis, vol. 1, p. 392 (date of volume, 1860).

White [Permian] limestone: Guadalupe Mountains.

Shell ovate, very gibbous, width and thickness about equal, one-third longer than wide in full-grown specimens; front subtruncate or slightly emarginate; sides rounded anteriorly and converging posteriorly to the beak at an angle of about 55°. Dorsal valve varying from circular-ovate to subcircular, convex, forming usually a regular curve from beak to front; old specimens marked with a slight mesial elevation in front and an obscure fold on either side, which becomes entirely obsolete before reaching the middle of the valve; cardinal edges rather deeply indented by the false area of the ventral valve. Ventral valve strongly convex, more elevated than the opposite valve; front marked with a shallow sinus, which usually becomes obsolete before reaching the middle of the valve; beak extended considerably beyond that of the dorsal valve, acute and strongly incurved; surface marked with fine concentric striæ of growth. Dimensions of a full-grown specimen: Length, 0.67; width, 0.52.

Formation and locality.—White limestone of the Guadalupe Mountains, Texas. It appears to be quite rare, only two specimens having been found.

This description, quoted from Shumard, does not altogether agree with any of the forms recently obtained from the Guadalupe Mountains. The species most nearly in agreement appears to be one which I have described as new, under the title of Dielasmina guadalupensis. A difference seems to exist in the plications, of which D. perinflata possessed in old specimens a well-marked mesial elevation on the dorsal valve and an obscure fold on either side, while the type specimen of D. guadalupensis has one well-marked and one obscure fold on either side of the mesial elevation. Other examples of D. guadalupensis lack the obscure plication, but there is in almost every case a plication on each side of the mesial fold which is well marked. Were it certain that no other difference existed, the name guadalupensis would not have been proposed, but a description without figures is so imperfect a manner of defining a species that in view of the confusion which a wrong identification in this case might introduce, it seemed unwise to make one which would be at least doubtful.

Genus NOTOTHYRIS Waagen.

This genus, hitherto unknown in North America, is even in the present instance somewhat doubtfully identified. The chief facts which have been ascertained bearing on its generic position (for I have not felt justified in sacrificing much of my very scanty material to this end) are these: The configuration impresses one as terebratuloid. There are a fairly distinct fold and sinus, which are not strongly elevated in themselves and in addition are somewhat masked by bearing a median sulcus and a median rib, respectively. The sinus and fold do not occupy the usual positions, but occur, the former on the dorsal and the latter on the ventral valve, respectively, so that the species must be considered as belonging to the antiplicatæ. The plications are few, rounded, and submarginal.

The shell structure is probably punctate, but is composed of minute fibers, which are finely fluted, thus conveying an impression of punctation which may be misleading.

Of the internal structures but little definite is known, for the specimens which show the interior are so imperfect that their specific relations are not quite certain. In the identifiable specimens the ventral valve appears to be without dental plates and there is little doubt that if present they would show through the exfoliated shell. The dorsal valve also appears to be without septa, but in this case their absence can not be so surely predicated as a probability. A silicified specimen referred to N. schuchertensis var. ovata, but not determinable with certainty, shows the absence not only of dental and septal plates in the ventral and dorsal valves, respectively, but the absence of a hinge plate in the latter, the short discontinuous (possibly broken) crura springing immediately from the sides of the shell near the hinge. This is rather against a reference to Notothyris, but is perhaps to be regarded as an abnormal or accidental feature. Another example, a silicified dorsal valve, specifically belonging to a very different group from the foregoing, has the normal Notothyris structure. Septal plates are absent, but there is a well-developed hinge plate pierced at its upper extremity by a rounded foramen.

It is almost certain that the type of structure found in N. schuchertensis can not be placed with the rhynchonelloid genus Terebratuloidea, which it resembles in some respects, because of the rounded submarginal plications and the presence of the fold and sinus in a reverse relation to the valves. Much more essential is the connection with the terebratuloids, especially with the genera Notothyris and Hemiptychina. The absence of dental plates, as it appears, distinguishes it from Dielasmina, and the absence of septal plates from Hemiptychina. Even if differences based on somewhat doubtful septal structures be eliminated, the character of the plications, especially as regards the fold and sinus, distinguish it from both the genera named, while not only this character, but also the internal structure, so far as known, ally it with Notothyris.

Two closely related varieties of this type have been recognized in the Capitan limestone of the Guadalupe section. They are allied to but not identical with Asiatic and European representatives of the genus. A third but very different type specifically has been obtained from the southern Delawares.

NOTOTHYRIS SCHUCHERTENSIS n. sp.

Pl. XV, figs. 25 to 25c.

Shell small, rotund, rather broadly ovate, with a straightened anterior outline. Ventral valve strongly convex, especially in the posterior portion. Beak large and much incurved. The surface is smooth, except toward the margin, where it is gently plicated. There is a low median fold, imperceptible except along the line of junction of the valves, and still further obscured by a faint median sulcus, so that the fold is surmounted by two plications. It is followed on each side by a sulcus, a lateral fold, and another slight sulcus. Both plications and sulci are rather strongly rounded.

Dorsal valve moderately curved. Along its anterior margin are a few plications corresponding to those on the opposite valve. These consist of a broad, low sinus with a central plication determined on each side by a rather high plication.

The internal characters have not been satisfactorily made out, but there are probably no dental plates in the ventral valve and possibly no plates in the dorsal. The shell structure is not foliaceous, as in Dielasma and most terebratuloids, but fibrous, yet at the same time appears to be punctate. The fibers seem to be finely wrinkled, conveying the appearance of punctation, even if this structure does not exist. Specimens of Pugnax swallowiana from the same horizon have a similar fibrous structure, but the fibers are not fluted or wrinkled. Furthermore, many internal molds from this horizon have a minutely papillose surface, due, it seems probable, to the obliquely fibrous structure of the shell. Many of these specimens are filled by crystalline calcite, and the appearance referred to may also be due to tiny crystals lining the inside. This appearance is very suggestive of a punctate shell structure. I am therefore doubtful whether the species under discussion really possesses this character. The form of the beak and foramen, however, is that of the terebratuloids and unlike that of any other group to which the species might reasonably be referred, except, perhaps, Terebratuloidea, and the configuration is against referring it to that genus.

This species and the variety ovata are related to N. simplex and N. inflata of the Salt Range fauna, to N. exilis and N. mediterranea as identified by Diener in the Himalayan region, to N. nucleolus of the Russian Gschelian fauna, and to N. [Rostranteris] inflata of Gemmellaro's Sicilian fauna, but I can not consider them specifically identical.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926).

NOTOTHYRIS SCHUCHERTENSIS var. OVATA n. var.

Pl. XV, figs. 26 to 26c.

This variety differs from the original species in being narrower and having the point of greatest width nearer the center of the shell. The plications are the same as in Notothyris schuchertensis. The median plication on the dorsal valve appears to be a little smaller than those lateral to it.

I have referred to this variety an imperfect specimen from the "dark limestone" (station 2930), the anterior portion of which is missing, so that the specific characters can not be satisfactorily determined; at least there is no apparent reason why it should not be placed with this species. Special mention, however, would not need to be given to this example save that it shows internal characters which seem to a certain extent to corroborate Waagen's more or less tentative diagnosis of the genus. Dental plates are entirely absent in the ventral valve. In the dorsal valve the loop is short and incomplete, but it may be broken. Septal plates seem to be absent in this valve also, and in fact a hinge plate of any sort if not absolutely wanting is rudimentary, the crura being attached directly to the two sides of the shell near the umbo.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966); middle of Capitan formation, Capitan Peak (station 2926); "dark limestone," Pine Spring (station 2930?), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

NOTOTHYRIS sp.

Pl. XXXI, fig. 7.

This name is introduced for an imperfect specimen which is interesting chiefly for the generic characters which it shows. It is a dorsal valve, and though a fragment it shows a well-marked median sinus. Consequently it belongs to the antiplicatæ. On the inside the apical portion is crossed by a well-developed hinge plate, which is not, however, supported by septa. Its posterior portion is perforated by a rounded or elliptical foramen, as described by Waagen. From its front margin project the crura, which are in this case broken off.

It seems fairly certain that this is a representative of Waagen's Notothyris, corroborating his description in a number of important particulars. In its specific relations it is very different from the two other Guadalupian types. It also presents wide differences, so far as can be determined, from any of Waagen's Indian species.

Horizon and locality.—Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

Genus HETERELASMA n. gen.

This group of shells, which appears to constitute an undescribed genus, is represented in the Guadalupian fauna by two species, of which Heterelasma shumardianum is taken as the type. This species has several rather striking peculiarities of configuration, such as the very compressed shape, which leaves only a narrow distance between the two valves, and the small appressed beak of the ventral valve; but as these features are less well marked in the second species, they would best not be included in the generic description. In regard to configuration, it can probably be said with safety, however, that, as in the Dielasmas, a more or less strong sinus was developed on the ventral valve and a corresponding fold on the dorsal, but in the case of Heterelasma a reverse plication is subsequently developed, more obvious on the dorsal than on the ventral valve, which produces an emargination of the front margin more or less pronounced, as the case may be.

On the interior the ventral valve bears two rather short dental plates, a feature very common among the Paleozoic terebratuloids, and in addition there appears to have been a low median septum. In the dorsal valve, however, a median septum is rather well developed. It is moderately high, and extends well forward. Septal plates are entirely absent, and even the hinge plate is rudimentary, being reduced to a thickening of the rim of the dorsal valve under the beak, at which the crura originate. The character of the loop is not known.

The shell substance has of course the usual punctate and foliaceous structure.

The configuration, and especially the internal structure, of these shells makes it impossible to refer them to any of the known genera, at least of Paleozoic forms, and in fact I do not know of any to which they are very closely allied.

HETERELASMA SHUMARDIANUM n. sp.

Pl. XV, figs. 21 to 22b; Pl. XXIX, fig. 10?.

Shell rather small,a flat, subpentagonal. Ventral valve moderately curved longitudinally, nearly flat transversely. Greatest width about three-fourths the length back from the front margin. Posterior lateral slopes gently curved, meeting at an angle of approximately 90°. Sides gently curved, slightly converging toward the front. Anterior outline strongly emarginate. The beak is small, wide, and incurved; flattened on the foraminal side into hooded expansions. The transverse curvature is slightly concave except near the posterior end. There is the merest suggestion of a mesial fold toward the front.

---

aA subsequent addition to our collection shows that, in addition to being relatively broad and flat, like the type specimen, this species may be narrower and much thicker.

---

The shape of the dorsal valve is like that of the ventral, but with a more obtuse posterior angle. The longitudinal outline is nearly straight; the transverse strongly convex. Beak small and prominent. Near the posterior end the transverse curvature is angular and gablelike, but this elevation is truncated below by a mesial sinus of increasing strength, which is deep at the front. A reentrant angle in the anterior outline is thus produced, on either side of which are projections made by the two plications.

On the interior there are two short dental plates in the ventral valve, with a long median thickening, like a low septum. The dorsal valve also has a median septum.

This species evidently resembles Dielasma problematicum Waagen, from which, however, it is at once distinguished by the absence of a fold in the ventral valve, as well as by the general configuration. Waagen refers his species to Dielasma. Gemmellaro refers to Hemiptychina species, which have a close resemblance to this. It is evident that H. shumardianum, since it possesses dental plates, can not be a Hemiptychina.

Associated with H. shumardianum at station 2926 were found a few small scalelike shells having a punctate structure and other characters indicating more or less close specific relationship. I have provisionally included them in the same species, as they have the expression which H. shumardianum would probably have presented in its younger stages. The immaturity of these shells is suggested by their slight convexity, as well as by the beak, which seems to be unusually erect for the genus. They have been placed with H. shumardianum in preference to H. venustulum because of their slight convexity. The fact that the two specimens found are of one size and that no examples intermediate between them and mature H. shumardianum have been found brings this procedure into some question.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2966); middle of Capitan formation, Capitan Peak (station 2926), Guadalupe Mountains, Texas. Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

HETERELASMA VENUSTULUM n. sp.

Pl. XV, figs. 23 to 24b.

Shell small, convex, subpentagonal. Greatest width at about the mid-length. Ventral valve gently flexed longitudinally. The posterior portion is rather strongly convex, the anterior half flattened. Beak of medium size, much incurved.

Dorsal valve moderately curved longitudinally, strongly curved transversely. No fold distinct from the subangular shape of the valve as a whole. At the front, and very restricted in its extension, is a well-marked median depression, giving rise to two short, low plications.

The form is related to H. shumardianum, and, like it, possesses in the ventral valve, besides the two dental lamellæ, a long median septum, which can be traced almost to the front margin. The dorsal valve has, likewise, a short median septum, but whether it is connected with other plates, as in Dielasma, I have not been able to ascertain.

Horizon and locality.—Middle of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (station 2926). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

Family SPIRIFERIDÆ King.

In scarcely any group of brachiopods is the difference between the Guadalupian fauna and that of the Salt Range more pronounced than in the family of Spiriferidæ. The generic representation is pretty much the same, both faunas possessing members of Spirifer, Spiriferina, Martinia, and Squamularia; but the Guadalupian fauna has Ambocœlia and the Productus limestone fauna Martiniopsis, not found in the other.

The eight species of Spirifer known from the Productus limestone are divided by Waagen into five groups. The group of S. striatus has no representatives in the Guadalupe Mountains so far as I am aware. Waagen refers to it two species, S. striatus and S. marcoui, remarking of the latter that it is a characteristic species of the American "Coal Measures." In this he is quite in error, for so far as my experience goes it is entirely absent from the "Coal Measures" of central and eastern United States, where most of our paleontologic work has been done. It is in fact a distinctly western type, which occurs in the Hueco formation below the Guadalupian. The group of S. tegulatus, also embracing two species, is present in the Guadalupian in the form or group of forms which I have called Spirifer sp. b, represented in our collection by but a few fragmentary specimens. The group of S. duplicicosta, to which S. wynnei of the Productus limestone belongs, is probably represented in my fauna by S. mexicanus and its two varieties and by Spirifer sp. a. There are, however, strong differences in the configuration of S. mexicanus, which make its relationship to S. wynnei rather remote, while some of the forms which I have placed in the same group with S. mexicanus are still less similar. Spirifer oldhamianus, S. alatus, and S. niger of the Indian fauna have, so far as known, no allied types in that of the Guadalupe Mountains, just as S. sulcifer of Shumard has none in the Productus limestone. On the whole the Spirifers of the Productus limestone are much more like those of the Hueconian than of the Guadalupian. In this entire section, however, including both its divisions, this genus is rather sparingly developed, in point of differentiation as well as abundance. In both particulars the Alaskan Carboniferous faunas are more fortunate.

Of Martinia Waagen distinguishes five species. The groups of Martinia glabra and M. corculum, each represented by a single species, appear to be absent from the Guadalupian fauna; but the group of M. warthii, comprising the three remaining Salt Range species, will perhaps embrace the only two types at present known from the Guadalupe Mountains, though the resemblance between the Guadalupian and the Indian species is not very great, considering how small the limits of variation really are within the group.

As would be expected, the Squamularias of the Guadalupian fauna resemble those of the Salt Range somewhat closely. I have contented myself with recognizing but three varieties in the Guadalupian. Waagen discriminates three species from the Productus limestone, and apparently his Reticularia lineata and R. elegantula correspond to the Guadalupian Squamularia guadalupensis and R. indica to S. guadalupensis var. subquadrata without any Indian equivalent for S. guadalupensis var. ovalis, but there may be intrinsic differences of sculpture, etc., which contradict the resemblances observable in configuration.

Between the Spiriferinas of the Indian and the American faunas I find it difficult to make a satisfactory comparison. The minute characters of sculpture, not easily described or represented by figures and very apt to be lost or obscured by preservation, have been destroyed in some of my Guadalupian shells and apparently in some of the Indian specimens also. I regard these as of much importance in discriminating and grouping species, and consequently will not venture to make extensive nor indeed more than tentative comparisons. In the Guadalupian fauna Spiriferina billingsi and its allies seem to form a rather well-defined group. These may be represented in the Salt Range by Waagen's group of S. insculpta, to which S. ornata is referred. This species appears to be very closely related to S. billingsi itself. S. evax, which I have placed in the same group with S. billingsi, is suggestive of the Indian shell which Waagen identifies as S. multiplicata Sow. Waagen does not place S. multiplicata in the same group as S. ornata, but with S. cristata, and if this disposition of the species is correct S. multiplicata and S. evax are not so similar as at first appears. In any event the type of S. billingsi apparently shows a wider range of variation in the Guadalupian fauna than has been found in India. S. cristata, as identified in the Indian fauna, is perhaps represented by the imperfectly known S. hilli and the other Guadalupian forms which appear to be allied to it. Perhaps all three species which Waagen assigns to the group of S. lima, viz, S. cristata, S. multiplicata, and S. nasuta, have here their only Guadalupian allies. In spite of its strongly lamellose surface, I believe that S. nasuta Waagen is not allied to S. billingsi, because it has a median plication in the ventral sinus and very coarse punctation. Perhaps its closest Guadalupian type is S. hilli and related species. S. vercheri of Waagen probably has no allied form in the Guadalupian, for he refers it to the group of S. transversa McChes and S. transversa has none. On the other hand, S. pyramidalis, S. laxa, and S. welleri seem to be without representatives in the Salt Range faunas.

In his paper on the Carboniferous faunas of Kashmir and Spiti, Diener distinguishes among the Spiriferidæ the four genera Spiriferina, Spirifer, Martiniopsis, and Syringothyris, the latter belonging probably to a different and earlier fauna. Martiniopsis, as I have already had occasion to remark, is not found in the Guadalupian, which contains the types Martinia, Squamularia, and Ambocœlia, not recognized by Diener. The only species of Spiriferina is one cited as Spiriferina cf. kentuckyensis, and while the Himalayan form is without much doubt distinct from our Pennsylvanian one, it represents a type which appears to be absent among the numerous species in the Guadalupian fauna. Thus in point of the genus Spiriferina the faunas are widely different, so far as known. Only slightly less different are the Spirifers of which Diener cites ten species. The group of S. fasciger, comprising three species, is probably represented by the rare and imperfectly known species or group of species which I have cited merely as Spirifer sp. b. The group of S. trigonalis, that of S. pinguis, and that of S. alatus, each containing one species, and the group of S. clarkei, with two, have no related forms in the Guadalupian fauna so far as known. Spirifer rajah and Spirifer undet. aff. rajah, constituting the group of Spirifer rajah, show considerable diversity of character. Some of Diener's figures seem to represent a species related to S. sulcifer Shumard, while others are in a measure comparable with S. mexicanus Shumard, though the resemblance is somewhat distant. With this possible exception the group of S. mexicanus the most abundant and characteristic Spirifer of the Guadalupian, does not occur in Diener's faunas, while most of his Spirifers are either absent from the Guadalupian or, in part, doubtfully represented by rare and little known species.

In Diener's second paper on Spiti fossils the Spiriferidæ from the lower beds appear under the titles Spirifer cf. strangwaysi, S. curzoni, Spirifer sp. ind. ex aff. Spirifer curzoni, Spirifer (Ambocœlia?) sp. ind. aff. fusiformis, and Martiniopsis cf. subpentagonalis. None of these species seems to bear much resemblance to any in the Guadalupian, and, in fact, the fauna to which they belong is probably a much older one. Spirifer curzoni has proved to be a Syringothyris. The species which Diener thinks is probably closely related to S. curzoni is almost certainly very different, as it has plications on the fold and is compared to the Devonian form S. disunctus var. sulcifer. The Spirifer related to S. fusiformis, to judge by the configuration shown by his figures, is probably not an Ambocœlia. It is somewhat surprising to find the genus Martiniopsis associated with a fauna which is probably either very late Devonian or early Carboniferous in age. Although the punctation of the shell might possibly be an appearance due to the bases of old spines, the form in question can hardly be a Reticularia, as one might surmise, because it has two dental plates in the ventral valve, while Reticularia has in addition a median septum.

The Spiriferidæ of the upper fauna of Spiti, according to Diener, are confined to five species of Spirifer, the other genera found in the Guadalupian being, so far as known, absent. The shells referred to Spirifer distefanii much resemble in their configuration Spiriferina evax of this report, but I have no species of Spirifer which is like them. If S. marcoui, S. fasciger, and S. nitiensis have any related species in the Guadalupian, it is in the imperfectly known Spirifer sp. b. The resemblance between S. rajah and S. sulcifer Shumard has already been commented upon.

Among the Spiriferidæ obtained in the limestone crag of Chitichun, Diener cites Spiriferina, Spirifer, Martinia, and Squamularia, all of them Guadalupian genera, which include, also, Ambocœlia. The only Spiriferina is identified as S. cristata var. octoplicata Sow., and it seems to be similar in configuration to S. billingsi, described by Shumard, though concentric ornamentation is said to be absent. No further account of the sculpture is given, and as this feature accordingly appears to be unknown it is impossible effectually to compare the Himalayan form with those from the Guadalupe Mountains.

A rather strong and unexpected resemblance between the Spirifers of the two regions appears to exist. S. musakheylensis, if represented in the Guadalupian, is represented in the very fragmentary material which I have called Spirifer sp. b. Spirifer wynnei, however, is closely allied to S. mexicanus, while S. tibetanus is very similar to S. mexicanus var. compactus, S. mexicanus var., and possibly also to S. sulcifer.

Diener distinguishes six species of Martinia in the Chitichun fauna, several of which are closely related to the two Guadalupian species. M. elegans and M. acutimarginalis correspond to M. rhomboidalis and M. shumardiana, respectively; but the less elongate types, such as M. semiplana and M. contracta, have not been found as yet in the Guadalupian fauna. The last-named species was originally described from the Mississippian of the United States, the genus not being known, in fact, in the Pennsylvanian of the Mississippi Valley and Appalachian region. White's identification of M. contracta in the upper Carboniferous of Nevada is based on a species of Squamularia. I must call attention in this connection to one of Diener's figures referred to Martinia elegans.a It has a different shape from the other specimen figured on the same plate, though not so different from the specimen figured on his Pl. IX. It hardly seems likely that Diener has fallen into this error, but the figure, in its configuration, sculpture, and even in its shell punctation, certainly is very suggestive of a large orthoid, such as Schizophoria or Orthotichia.