Geological Survey Professional Paper 58 The Guadalupian Fauna

SPONGIÆ.

The sponges of the Guadalupian form a remarkable feature of the fauna, both in abundance, variety, and peculiar development. This group is apt to be neglected, in the field as well as in the laboratory, and in the present case no special effort was made when collecting to obtain material. In fact, attention was particularly centered on the brachiopods and mollusks; yet the sponges have proved not the least novel and varied element of the fauna.

The preservation of my specimens leaves in many cases much to be desired, and often it has been impossible to refer them even to the genus with certainty. The two types of Silicispongiæ and Calcispongiæ have been recognized, not because of any difference in their present mineral composition, but by reason of apparent structural affinities. In the white limestone of the Guadalupe Mountains, where sponges are especially abundant, both types lack the original siliceous composition which one of them is supposed to have possessed. It is chiefly at this horizon alone that the Calcispongiæ, so far as known, are found. In the lower beds of the Guadalupe section and at some of the outlying localities the Silicispongiæ still retain their siliceous nature, though probably not the original siliceous material or form.

Owing to the aberrant character of some of the organisms here referred to the sponges, their determination as belonging to the Lithistida of the Silicispongiæ or to the Calcispongiæ is still debatable. Steinmannia, Amblysiphonella, etc., which seem by common consent to be grouped with the calcareous sponges, have completely lost the spicular structure of their walls, and the types from the Capitan in which this has happened, especially those that manifested an appreciable analogy in their organization to the genera named, have been placed with the same group, as indicating apparently the best disposition which it is at present possible to make of them.

Owing partly no doubt to their real scarcity, but in some degree probably to their unattractive character and the little promise of return which they hold out, sponges have been described in but few of the works with which it seemed desirable to compare the Guadalupian fauna. One of the exceptions is, of course, Waagen's monograph on the faunas of the Salt Range of India. In this work species belonging to the genera Amblysiphonella and Steinmannia are described, showing an agreement, so far as they go, with the Guadalupian fauna, which possesses, however, a much greater abundance and variety of these organisms. In his paper on the Chitichun fauna No. 1 Diener cites a species of Amblysiphonella and a problematicum which may prove to be one of the Calcispongiæ, though of a genus different from any of the Guadalupian types and probably new.

Among the scattered references in which sponges of this period are treated, perhaps the most noteworthy, aside from that of Waagen, are several which deal with the European Permian. In the Russian Carboniferous series these organisms would appear to be very rare. Netschajew, however, in his account of the Permian faunas of eastern Russia figures two specimensa representing a type which is rather suggestive of some of the Guadalupian genera (Guadalupia cylindrica). Whether this resemblance is entirely superficial or has some real basis in structure can not be told.

---

aNetschajew, A., Kazan Obshchestvo Estestvo-Ispytatelei, Trudy, vol. 27, 1894, Pl. I, figs. 27, 36.

---

In his monograph on the Dyas, Geinitz cites a number of sponges, most of them, it is true, quoted from King. The two German species which he figures are too imperfectly known for me to ascertain their relationship to the Guadalupian forms.

King's monograph, already referred to, contains the citation of five species of sponges, which he distributes among the genera Tragos, Scyphia, and Mammillopora. The form figured as Scyphia tuberculata is very suggestive of that which in the Guadalupian has been named Cystothalamia nodulifera, but here again the means are not at hand for determining whether the resemblance is accidental or intrinsic. The four other species do not recall anything in the Guadalupian fauna, although the form referred to Mammillopora mammillaris may be related in point of structure, though less so in growth, to Virgula neptunia.

Hind's catalogue of sponges in the British Museum shows a rather limited list of species, and one which manifests very little resemblance to the Guadalupian. Among the monactinellids he cites one species of Reniera.? and one of Haphisteon; among the tetractinellids one species of Geodia and one of Pachasterella; among the lithistids one species of Doryderma; and among the hexactinellids two of Hyalostelia and four of Holasterella. Mortiera vertebralis, a type of unknown affinities, is also recorded. We note here the large development of hexactinellids and the absence of calcareous sponges, almost the reverse of what is found in the Guadalupian.

I should not neglect to mention in this connection the fauna collected by Barois from Sebargas, Spain, in which Steinmann described the genera Sebargasia, Amblysiphonella, and Sollasia. I do not know what the associated fossils were, nor the age they indicate, but in the abundance of sponges, and to a certain extent in the character of those present, though probably but one of the genera found in Spain is represented in the Guadalupian, a certain affinity is shown with the fauna of the Guadalupe Mountains.

Lastly, in North America we have five genera of sponges in the Pennsylvanian, only one of which—Amblysiphonella—is common to the Guadalupian.

Lacking in conclusiveness as these comparisons probably are, they certainly show to some degree the unusually prominent place these organisms take in the Guadalupian fauna, and the unique structural types by which that fauna is distinguished.

SILICISPONGIÆ.

Order LITHISTIDA.

Suborder TETRACLADINA.

In the Guadalupian fauna three generic types are referable to the lithistid sponges (Anthracosycon, Virgula, and Pseudovirgula), and according to the best of my present knowledge they have been placed with the Tetracladina. The intimate spicular character and construction of these sponges has not been determined with ease or certainty, for the spicules are small and so completely consolidated that it is practically impossible to tell where one begins and the other ends.

While these genera have been placed with some confidence among the Lithistida, and less confidently among the Tetracladina, I have avoided entering on the subject of assembling them into families, not only because of uncertainty or incompleteness of information, as above recorded, but also from a doubt as to what should constitute a family among these forms. It appears, at all events, that Anthracosycon should be referred to a different family from Virgula and Pseudovirgula, and possibly these two genera also should be placed in different families, a decision depending on the presence of certain characters which are as yet doubtful.

Genus ANTHRACOSYCON n. gen.

Shape turbinate or pyriform, attached by the smaller end, without a peduncle or with but a small, ill-defined one. Cloaca represented by a slight depression on the upper surface, from which descend several (three or four) tubular openings through the axial region of the sponge. Ostia small and numerous. The spicular structure consists of minute, regular tetraxons, more or less digitate toward the extremities of the arms (?), where they unite with one another to form a rigid skeleton. Loose monaxial spicules associated with typical specimens may belong to the same genus.

Not having found any Carboniferous genus to which this sponge could consistently be referred, I have erected a new one for it. Its systematic position seems to be clearly with the Lithistida, and I would be disposed to place it among the Tetracladina.

Type.—Anthracosycon ficus.

ANTHRACOSYCON FICUS n. sp.

Pl. XXIV, figs. 1, 1a, and 2.

Shape turbinate or ficiform, more or less oblique, attached at the small end. Peduncle small or absent. Cloacal depression slight. Ostia small, averaging about one-half millimeter across, circular, closely arranged at distances from one another of about their own diameter. They tend to be connected, especially in the upper portion, by channels whose general direction is radial from the cloaca.

The spicules are very small, apparently regular tetraxons, more or less branching at their ends, by which they are cemented to form a rigid skeleton. With these sponges are associated numerous loose monaxial spicules, which, if they belong to the same organism, probably had a dermal position.

The type specimen of this species, which was obtained in the black limestone south of El Capitan, is of medium size, having an axial length of 18 mm. and a greatest diameter of 24-mm. It is strongly oblique and much flattened above, with a scarcely perceptible cloacal depression. There are slight constrictions of growth parallel to the upper surface.

Horizon and locality.—Basal black limestone, Guadalupe Point, Texas (station 2920).

ANTHRACOSYCON FICUS var. CAPITANENSE n. var.

Pl. VII, fig. 10.

When I first described the species Anthracosycon ficus, with the typical specimen, which was obtained from the black limestone at the base of the section, was placed one from the white limestone of the Capitan. The two are very similar in external form—much more nearly identical than would be expected from their wide separation in geologic horizon, the chief difference being that the typical specimen is strongly oblique and that from the Capitan more nearly erect. Another difference consists in the fact that the Capitan example, which is considerably smaller, has distinctly larger ostia, while more careful comparisons seem to show that the spicular structure is also coarser. Some latitude must be allowed to the last statement, for it is just possible, though I do not believe it to be the case, that in the typical A. ficus we have not the original spicular structure of the sponge, but a fibrous, siliceous mass retaining only the general arrangement.

On account of these differences of structure, for that manifested in the inclination of the axis is probably only an individual character, it seemed necessary to separate the Capitan specimen from those obtained at a lower horizon. Another specimen from the Capitan subsequently came to hand, much smaller than the first, which on being broken lengthwise shows a group of several relatively large tubes occupying the axial portion and apparently representing several cloacæ debouching in the depression which occupies the upper end of the sponge.

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

ANTHRACOSYCON? sp.

Associated with portions of a sponge which have been referred to Anthracosycon ficus was found a little group of consolidated spicules, distinguished from them by having a distinctly coarser mesh. On more careful examination the component spicules appear to be of the four-rayed type and to have the arms, or at all events three of them, considerably branched. It did not, therefore, seem appropriate to place this form with Anthracosycon ficus because of the larger sized and more conspicuously branched spicules, and even the assignment to the same genus is attended with much uncertainty, since the construction of the sponge body as a whole is yet unknown. More exact knowledge as to the character of the spicular elements, as well as of the general structure, will be needed before the affinities of this form can be determined.

Horizon and locality.—Basal black limestone, Guadalupe Mountains, Texas (station 2920).

Genus VIRGULA n. gen.

Sponges belonging to this genus are rather abundant in the Guadalupian and can be differentiated into several species. The spicular mesh is rigidly consolidated, and it is very difficult to determine with certainty the character of the typical individual spicules, since there is no definition between them. It is inferred to be a regular tetraxon, with the arms but little divided, if at all. From the distinct outline in rock sections of certain of these sponges a dermal layer of some sort may be inferred, but its character can not be determined from the material in hand.

The general shape in which these organisms grow is subramose, sometimes in relatively slender stalks with few branches, and at others irregularly and frequently branching. There is in some specimens a tubular cloaca extending part way through the center of the sponge, but this has not been demonstrated as a permanent feature. Ostia appear to be absent.

Type.—Virgula neptunia.

VIRGULA NEPTUNIA n. sp.

Pl. VII, figs. 11 and 12.

While this species is fairly abundant in the Capitan limestone, it so happens that in the most perfect specimens, those which best show the shape, the structure is obscured, while the specimens in which the spicular architecture is especially well preserved are fragments. The general shape seems to be more or less irregularly cylindrical, rather frequently bifurcating or putting out short branches. It seems very probable that in the upper ends of the branches there was a tubelike cloaca of greater or less depth, but this can not be affirmed positively. The branches sometimes attain a diameter of 10 mm. and rarely exceed it to any extent. Usually these bodies are seen on the broken surface of the dense limestone, nearly circular when the section is transverse, more or less elongate when otherwise directed, and irregular when near the branching point. When the preservation is good the outline is fairly distinct and entire, indicating, it would appear, that there was a thin dermal layer, the spicular structure of which, however, has not been made out.

Where well preserved the skeleton is seen to consist of a rigid complex of cemented spicules, which are of rather large size. This complex is not easily analyzed into its constituents elements. So seldom do four rays diverge from a point, and so often three, that it seems highly probable that the spicular unit was a tetraxon. Absolute regularity is not maintained in this respect, however, and one or two specimens, which appear to be for the most part composed of tetraxons, show part of the structure as if made up of continuous parallel rays with cross arms at right angles to them, a type of structure which is most naturally associated with the Hexactinellida. Nevertheless, I am fairly satisfied that this sponge belongs with the Lithistida.

These fossils are liable to be poorly preserved, and instead of appearing composed of distinct spicules the structure is sometimes represented merely by a fine mottling of opaque and semitransparent whitish dots, or the structure may be lost altogether, the definition of the sponge as a whole, nevertheless, remaining fairly distinct. When poorly preserved, it is of course difficult to distinguish this form from even such types as Guadalupia cylindrica and Cystothalamia nodulifera, which, though widely different in structure, have a similar growth, and it is possible that some of the obscure specimens placed under this name may be of a bryozoan nature; in their present condition it is impossible to be sure.

Horizon and locality.—Middle of Capitan formation, Capitan Peak (station 2926), and peak north of Pine Spring (station 2902?), Guadalupe Mountains, Texas.

VIRGULA RIGIDA n. sp.

Pl. VII, fig. 13.

I am a little uncertain as to the proper position of the specimen on which this subdivision is chiefly based. As will be seen from the figure, it consists of two straight cylindrical branches connected at their bases and marked by irregularities due to growth. The spicular network is largely obliterated, but seems to be the same as in V. neptunia. There is no cloaca. The external surface is seen to be reticulated more finely than a spicular framework like that of typical V. neptunia would produce. It may be the dermal layer. In a general way one notices that the dominating system of lines in this superficial network is longitudinal.

A number of other specimens have been assigned to this species, but the reference is more or less uncertain in proportion to their very imperfect preservation. They consist, so far as known, of simple cylindrical stems without bifurcations. Typical specimens of V. neptunia are short, often tapering, with a tendency to send out imperfect branches. On account chiefly of these differences in the mode of growth it seemed best to keep the present forms distinct from V. neptunia.

Horizon and locality.—Top of Capitan formation, Capitan Peak (station 2905?); middle of Capitan formation, Capitan Peak (station 2926), and peak north of Pine Spring (station 2902?), Guadalupe Mountains, Texas.

VIRGULA RIGIDA var. CONSTRICTA n. var.

Pl. VII, figs. 14 and 15.

This variety is based on some slender specimens which have a diameter of 7 mm. or less and are distinguished by somewhat contorted growth and by more or less pronounced constrictions, which are neither parallel to one another nor equal distances apart. These constrictions are purely external and are independent of structures within. They do not mark internal partitions, the spicular skeleton being uninterrupted by them.

The spicular structure is similar to that of V. neptunia. A deep tubelike cloaca seems to be a constant feature.

This form is distinguished from V. rigida by its smaller size and more pronounced constrictions.

Fig. 15 of Pl. VII represents a specimen which has been referred here with some doubt. Weathering has caused it to show clearly the reticulate surface, due most probably to spicular framework. In this particular it has a different appearance from the typical examples, which have been broken from unweathered rock. Aside from this and one or two trivial differences, the only other point of disagreement is the complete absence of a cloaca, a structure which is present in the typical examples.

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

Genus PSEUDOVIRGULA n. gen.

This term is introduced for a species which I originally placed with the genus Virgula, but which on reconsideration it seems desirable to refer to a distinct group. The general appearance, as cylindrical stemlike growths, is similar, and the character and arrangement of the spicules also resemble those of Virgula. On the other hand, Pseudovirgula tenuis is provided with large ostia, structures which have not been observed in the other group. Inside the dermal layer, between it and the consolidated spicular median portion, is an empty or hollow zone, which if a real character (this being very doubtful) would constitute an important distinction between the form in question and those subsumed under the title Virgula. Again, Pseudovirgula tenuis is somewhat obscurely divided off into structural segments. The presence of large ostia is the only difference subsisting between Pseudovirgula and Virgula which can perhaps be called real and constant, but in view of the indications of the additional differences mentioned it hardly seems a sound course to place both types in the same genus.

Type.—Pseudovirgula tenuis.

PSEUDOVIRGULA TENUIS n. sp.

Pl. VII, figs. 16, 16a, and 17.

This species forms small, more or less irregular, cylindrical branches, which probably sometimes divide. The branches seem to vary in diameter from 2 to 4 mm. The spicular skeleton is as in Virgula neptunia, but considerably finer. The spicules are arranged so that in tangential section one axis tends to form, with others similarly arranged, continuous, longitudinal lines. In cross section no particular arrangement is noticeable. Individual spicules can not be distinguished, so complete is the cementation of the framework, but as frequently three axes diverge from a point the typical spicular element is interpreted as being a tetraxon, though it is possibly a hexact. In the typical example the spicular structure does not continue quite to the sharply defined perimeter of the sponge body, but leaves a somewhat irregular zone, which would appear to have been hollow; I suspect, however, that the spicular framework actually did connect with the dermal layer and that its peripheral portions have been obliterated. In the typical specimen the outer surface is furnished with a number of relatively large, somewhat tubularly projecting mouths, or ostia, and in another specimen these structures are seen to penetrate the spicular net work as wall-less tubes. In one example in which a longitudinal section is seen, the sponge body appears to be obscurely divided off as if by several cessations and renewals of growth. Some specimens have the spicular mesh finer than in others.

In its general mode of growth and spicular structure this form resembles V. neptunia, so that one would at first be disposed to refer them to the same genus, but the presence of large ostia is a character not known in any species of Virgula. It is certainly not present in the form described as V. rigida, and could hardly fail of preservation if it had been present.

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

Order HEXACTINELLIDA.

Suborder DICTYONINA.

The hexactinellid sponges are represented in the Guadalupian fauna by a single species of so unusual a character that it clearly belongs to a new genus and probably to a new family. I have not, however, introduced a family designation for this form, awaiting more complete and exact data as to the proper character of its organization before essaying to fix its relations and affinities with others of the Hexactinellida.

Genus STROMATIDIUM n. gen.

This name is introduced for a new genus of siliceous sponges whose general shape and construction are as yet unknown, but whose spicular structure has been definitely ascertained. It is mainly made up of layers of spicular reticulations, separated and held together by pillarlike rays. The spicular elements are believed to belong to the hexactinellid system, one ray being aborted, one constituting the pillar ray, and the four others forming the mural layers. These mural rays apparently branch, probably several times, all of the elements lying in the same plane and inosculating with one another and with adjacent spicules so completely that the constituent arms can no longer be differentiated. The more or less regular mesh which results appears to be quite continuous and homogeneous. Associated with fragments having the foregoing structure are great numbers of loose spicules, which may belong to the same sponge, fulfilling the function of flesh or dermal spicules. A few of these are hexacts but most are monacts, some short, curved, and tapering to blunt points, others much longer and very slender, sometimes nearly if not quite straight.

While it will appear from what has been said that these sponge fragments probably belong to the Hexactinellida, the difficulty of discriminating individual spicules in the mural layers is such that the individual elements may possibly have been normal tetracts and the systematic position really with the Lithistida. If a hexactinellid, this is clearly an aberrant form, and while it probably belongs to the Dictyonina, is of doubtful family position.

Type.—Stromatidium typicale.

STROMATIDIUM TYPICALE n. sp.

Pl. XXVII, figs. 7, 8, 8a, 9, 10, and 10a.

The size and general shape of the entire sponge in this species are unknown, the parts which are preserved permitting only an adumbration of these characters. The nature and arrangement of the spicules, on which the species is accordingly based, are, however, well shown by the typical specimens. The spicular skeleton consists of more or less regular, superimposed layers, which are connected by pillars having a radial direction. These slender connecting pillars are evidently single rays of spicules, whose other rays lie in a plane normal to them and inosculate with one another to form the superimposed reticulated layers. The reticulation of the latter is so complete that it is impossible, in the specimens examined, to discriminate individual spicules. Spicular centers are often indicated by broken ends of the pillar rays, the initial number of mural rays proceeding from which is usually four, meeting rather regularly at angles of 90°; but these rays appear to branch and to inosculate with one another and with those of adjacent spicules to form a more or or less regular mesh, the apertures of which are approximately circular, but of variable sizes. The pillar rays are apt to be rather far apart, so that if some of them have not been destroyed without leaving any very perceptible traces, the mural rays may branch several times before meeting those of adjacent spicules.

I regard this structure as being made up of hexacts, the sixth ray in each case being aborted (for the pillars do not seem to be continuous from layer to layer) and four of the five other rays branching and inosculating to form the mural reticulations. This belief finds additional support in the discovery of loose normal hexactinellid spicules associated with this sponge, and of one of them in fact apparently forming an integral part of the skeleton. In this connection mention must be made of loose spicules, mostly monacts, which occur in really great numbers, associated with fragments of this sponge. Many of these are small, slender, slightly curved, and tapering to a point at both ends. Others are nearly or quite straight, very much longer, though equally slender, and either gently tapering or truncated, in the latter case probably being merely broken segments of long acicular forms. The spicules usually show clearly the fine central canal, a structure, the retention of which, taken in connection with the present siliceous composition of the spicules, suggests that the original material was also siliceous.

Considerable intervals, to speak relatively, are left among the sparse pillar rays and the spicular layers which they connect. This space may have been occupied by the loose spicules of which mention has already been made. One large hexact does in fact occupy some such position, as already noted, though its location in the sponge at all may be accidental.

Owing to the arrangement above described, the anastomosing mural rays form a much more firm and solid structure than their union with one another by means of the pillar rays, so that it is common to find thin scalelike fragments of the mural reticulation which retain but little evidence of the pillar rays that originally united them. These fragments are usually more or less strongly curved, indicating that aside from irregularities in the layers themselves the shape was probably more spherical than planate. Doubtless a canal system, of which no trace is found in the small pieces thus far examined, was originally present, but its character is unknown.

Another feature of this sponge which must not be overlooked consists in the development of numerous small spines or papillæ upon the mural rays. These occur in varying degrees of eminence and are usually more striking on one side than on the other. In one example, one side of which is apparently smooth, the other side, as a result, it would seem, of an unusual development of these apophyses, is covered by a complete entanglement of fine branching structures like an adherent layer of delicate inosculating spicules. This is interpreted as being part of the same sponge to which the other fragments belong, representing perhaps some specialized portion of the anatomy. In a corresponding manner the pillar rays also, though to a limited extent, bear spinelike developments.

Horizon and locality.—Delaware Mountain formation, Guadalupe Point, Guadalupe Mountains, Texas (station 2963?). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2969).

CALCISPONGIÆ.

Order SYCONES.

The sponges of the Guadalupian fauna comprise some very remarkable types referable to the Calcispongiæ. First in interest among these is the genus Guadalupia, which embraces forms so peculiar that it has seemed necessary to regard them as representing not only a genus but a family which is new. Scarcely less remarkable is the genus Polysiphon for which also the establishment of a new family seemed to be demanded. Standing less aloof from types previously described is Cystothalamia, a group which though more obviously related to existing genera I have nevertheless felt compelled to regard also as a new genus and family. To the previously described genus Amblysiphonella, representing the Sphærosiphoniidæ, the Guadalupian furnishes one new species. The Spha&elig;roceliidæ also are represented in the Guadalupian by species in one instance belonging to the genus Steinmannia and in another referred, with much hesitation, to Sollasia. These five families, with their six genera, I am at present placing in the order Sycones.

Family GUADALUPIIDÆ n. fam.

Although several species have been discriminated as belonging to the genus Guadalupia, I find much difficulty in framing a description of this new family, because since only one genus is known it is impossible to separate family characters from those which are solely generic. The distinctive features suggested by the species of Guadalupia are the growth, usually in lamellar expansions, and the construction of the walls, which are composed of tubes having a direction normal to the two surfaces, the superficial layers being reticulated and apparently formed of large, mutually consolidated spicules. A cloaca, strictly so called, seems to be absent, though possibly the whole organism may be analyzed into a colony of cylindrical individuals having some of the characters of the Sphærosiphoniidæ. In this case the cavities of the tubes would be cloacæ. The surfaces, while porous, probably have nothing corresponding to ostia.

Genus GUADALUPIA n. gen.

This generic name is introduced for a structural type which is fairly common in the white limestone of the Guadalupe Mountains, but while a number of specimens have been obtained they are so liable to be fragmentary and to have the more minute structure obscured that their study has been pursued under some difficulty and attended with incomplete success. These organisms are believed to belong to the class of Spongiæ, but they are sufficiently obscure and aberrant to make their exact zoological position a matter of some doubt. They assume a variety of shapes—massive, cylindrical, frondlike—the walls being of considerable thickness and of unusual construction. The most striking feature, and one which is most largely developed, occupying the greater portion of the mural body, consists of a series of cylindrical tubes parallel to one another and perpendicular to the two walls which they connect. These tubes are rather closely arranged, usually almost in contact. Their walls are substantial, yet at the same time a considerable caliber is left within. They contract somewhat at either end, and are gently curved. They are also intercepted at very variable intervals by straight, flat diaphragms, which may be close together or, on the other hand, almost absent, though the continuity of the tube is generally more or less interrupted close to the point where it terminates. The structure of the surfaces in which the mural tubes terminate has not been ascertained in its details. It consists of a rather open but moderately fine mesh, which is probably composed of large spicules, the exact shape of which it has not been possible to make out. It is not certain that both walls had the same structure, though such is not improbably the case.

Various shapes are assumed by the organisms, which possess, in the main, this sort of structure, but it is probable that they may have had initial stages very much alike. Such a condition appears to be presented by the young specimen figured on Pl. VI, which is attached below by a thickened and rather dense basal portion. Its growth has begun to form an explanate shape, the tubes being perpendicular to the plane of expansion. The upper wall is thin and has a reticulate structure. This specimen is provisionally referred to G. zitteliana, which in its mature stages is a branching frond, and it is evident that the mode of growth in which this young example started to develop would have to be changed in order to produce the configuration of the mature sponge. Similar modifications of growth must be supposed if the mature shape is that of G. cylindrica.

The spicular structure of these organisms has been to a considerable extent obscured. The dermal layer, as already remarked, seems in many cases quite clearly to be the result of large consolidated spicules, while the mural tubes frequently appear to be entirely structureless. It seems probable, however, that they also are made up of interlocking spicules of large size, for in some thin sections their walls are seen to be marked with rather regularly distributed dots, which probably represent sections through the arms of spicules. At the same time, these darkened spots in the walls of the mural tubes have, so far as observed, always a circular shape, while it would be expected that in some cases at least the direction of the section would coincide with the longer axis of the spicular rays.

In some respects the structure of this sponge suggests the geologically much older genus Receptaculites, and one might be tempted to interpret the mural tubes as the axes of large spicules, but the analogy soon ends, as there is nothing to correspond to the other rays of such spicules, while the occurrence of transverse partitions dividing the tubes entirely negatives the interpretation of them as spicules.

Type.—Guadalupia zitteliana.

GUADALUPIA ZITTELIANA n. sp.

Pl. VI, figs. 1 to 1d, and 2 to 2b.

This species occurs in the form of broad, gently convex fronds, which, as in the typical specimen, are sometimes branching. One of the largest fragments referred to this species is about 45 mm. square. The thickness seldom equals 10 mm. and averages perhaps but 6 or 7 mm. It diminishes somewhat toward the margins of the frond, which are rounded. The fronds are marked with transverse irregularities and wrinkles, which presumably represent stages of growth. The mural tubes average 1 mm. in diameter, or perhaps a little less, and the partitions by which they are crossed vary much in frequency in different cases. The dermal mesh is much finer than the tubes, but at the same time is rather coarse and made up of large spicular elements.

A small specimen which appears to belong with this species has a small peduncular attachment, the main portion of the organism being explanate, with a flattened upper surface, to which the mural tubes are perpendicularly directed. While the growth in this specimen is approximately symmetrical, it is to be supposed that one side would have expanded at the expense of the other, to produce the frondlike shape which distinguishes mature forms like the type specimen.

It is a little surprising that this species has a convex instead of a flat shape, since the first inference is that it is represented by fragments of what was originally a cone or a cylinder, but the structure is so finished at the margins as to furnish evidence that this was not the case.

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

GUADALUPIA ZITTELIANA var.

In this form, which is known only from limited and imperfect material, the mode of growth is in flat lamellar expansions having a thickness of 4 mm. or a little more. The mural tubes come about eight or nine in a distance of 5 mm. They are, therefore, considerably smaller than in typical G. zitteliana, and this constitutes at present the chief reason for distinguishing the two forms. It is possible that this is the same species as G. zitteliana, by reason of being not younger or undeveloped portions of a frond, but merely a more delicate variety connected by intermediate stages not yet discovered. On the other hand, it is possible that with the small proportions are associated other differences which in my imperfect material it is impossible to make out.

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

GUADALUPIA CYLINDRICA n. sp.

Pl. VI, figs. 3 to 3c.

The type specimen of this species has a somewhat cylindrical shape, tapering gradually at one end, and with the cross section approximately circular. The largest diameter is about 13 mm. and the length is estimated at 40 mm. The outer surface is obscured, but other specimens referred to the same species are marked by gentle swellings and constrictions, due to irregularities of growth.

Guadalupia cylindrica not only has a cylindrical form, but is hollow, the internal cavity not improbably corresponding to the cloaca of other sponges. In the typical example the thickness of the walls is about 3 mm. The mural tubes have essentially the same size as in G. zitteliana, about five or six occurring in a linear distance of 5 mm., and their direction is of course normal to the two surfaces which they connect. They are intercepted by tabular partitions, as in G. zitteliana, and, as in that species, the dermal layer is composed of large spicules whose character has not yet been determined. The walls of the mural tubes sometimes show dark spots, which appear to represent sections through spicular rays, and from this it may perhaps be inferred that the tubes also were composed of consolidated spicules, as in the dermal layer, though in most sections the tube walls appear to be structureless, even when the spicular mesh is obvious in the latter.

It is only in a sense that this species can be said to be hollow. In the type specimen the mural tubes terminate some distance before reaching the axis, their ends forming a rather irregular inner boundary whose ragged outline contrasts with the regular exterior one, but what would otherwise be a hollow cavity is partially filled by cysts. (See figs. 3 and 3a of Pl. VI.) Whether these form part of the real sponge body, or their presence is adventitious, I have been unable to determine, but they have in a general way some of the structural features of the sponge with which they are associated. Some of the specimens referred here appear to be without the central cavity, but as it was probably confined to the upper portion of the organism it is not necessary to suppose that they were entirely without it.

All the specimens at present referred to this species are single stalks, which have not divided to form branches; but one individual shows on its upper margin what seems to be the inception of a small bud, although it is possible that it is the commencement of a new and alien individual.

Nine or ten specimens have been referred to this species, and they show considerable diversity in appearance, but this is due in part to the alteration which has to a greater or less degree affected all the Guadalupian sponges and made it extremely difficult to identify many of them even generically.

Horizon and locality.—Top of Capitan formation, Capitan Peak (stations 2905 and 2966); middle of Capitan formation, Capitan Peak (station 2926) and peak north of Pine Spring (station 2902), Guadalupe Mountains, Texas.

GUADALUPIA CYLINDRICA var. CONCRETA n. var.

Pl. VI, figs. 4 to 4b.

The specimen which has been separated under this title appears to have arisen not so much by gemmation as from a natural lateral expansion, or possibly from the amalgamation of two or three separate individuals which have independent but contiguous points of origin, and since they developed in contact became organically confluent. On the upper portion, at all events, there are separate "cloacæ" and the mural tubes are directed in a measure toward independent centers. This is only to a certain extent true of the specimen, for though at one end there appears to be a circular "cloaca" of about the size and character of that in Guadalupia cylindrica, followed laterally by two other similar but much smaller ones, for the rest the growth seems to have assumed a bifoliate arrangement, without any central opening at all. If this colony originated by gemmation from a single original individual the budding or fission took place at a very early stage.

In other respects this form is closely allied to the type specimen of G. cylindrica, having an external spicular layer and mural tubes of about the same diameter. It is possible that it may have arisen from a typical example of G. cylindrica by some unusual process of increase.

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

GUADALUPIA CYLINDRICA var. ROBUSTA n. var.

Pl. V, fig. 12.

Several specimens are subsumed under this title. Their preservation is very unsatisfactory, but they seem to have the essential structures of Guadalupia cylindrica. At the same time the epidermal spicular layer, though presumably present, has not been observed. The size is considerably greater. The specimen figured is branched, but the internal structures have been obscured, so that intrinsically one can not determine whether it actually belongs to Guadalupia or not. Another example shows the characteristic mural tubes, but the central cavity seems to be lined with a relatively thick, dense layer which the mural tubes do not penetrate. In another part the same specimen appears to be filled up centrally by this secretion, a "cloaca" being absent. A third example has a diameter of 33 mm. and remarkably large mural tubes. The "cloaca" appears to be open, and I am not prepared to affirm definitely the presence or absence of an inner layer. The specimen is re-referred here with doubt.

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

GUADALUPIA FAVOSA n. sp.

Pl. VII, fig. 9.

This species is based on a small, apparently massive, somewhat irregularly shaped specimen, whose greatest diameter is about 24 mm. The mural tubes have a small diameter, and while approximately circular are in close contact. About six or seven occur in a linear distance of 5 mm. I am not sure that the dermal layer has been observed, but what appears to be this portion is full of small openings, more or less uniform in size and regular in distribution, which may be circumscribed by the arms of large-sized spicules. This layer is rather thin and the pores are small, being about one-fourth the diameter of the tubes.

In this specimen the walls of the tubes show a singular structure not noted elsewhere. They are now represented by dolomite (?), but this material has the appearance of being very finely porous, resembling in appearance the shells of some punctate brachiopods. This structure is probably rather secondary than organic.

The irregular massive growth and the small size of the mural tubes distinguish this form from the others here described.

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

GUADALUPIA DIGITATA n. sp.

Pl. V, fig. 13.

This name is proposed for a specimen which is rather regularly cylindrical in the lower part, where it has a diameter of about 11 mm., gradually expands above, and is terminated by an umbel of rather stout, equal branches. Four of these are developed on the side of the specimen which is exposed to view, and if symmetrical there must have been about eight in all. The lower part of the body is solid, while the branches have cylindrical "cloacæ"

The structure in this specimen is largely obscured, but the branches show not only the "cloacæ" but traces of mural tubes. The main part of the stem, moreover, is covered with regular little elevations or monticules, which may have had the same origin. These are considerably smaller than the mural tubes in other species of the genus, and also smaller than the mural tubes in the branches of the same specimen. They might be thought to represent the inner diameters of the tubes, but are too closely arranged to be so interpreted.

There might at first seem to be some question as to whether the digitate end were not really the basal end, and the branches really roots. The fact, however, that the branches are clearly hollow—i. e., occupied by "cloacæ," while the opposite end is solid, seems adverse to such a view, while other members of the genus are known to attach themselves merely by cementation, without developing rootlike processes.

Horizon and locality.—Middle of Capitan formation, peak north of Pine Spring, Guadalupe Mountains, Texas (station 2902).

GUADALUPIA sp.

Pl. VI, fig. 5.

Fig. 5 of Pl. VI represents a specimen, probably belonging to the genus Guadalupia, whose specific position at least is uncertain. It appears to have been an elongate, generally cylindrical body composed of tubes having an approximately radial direction. In cross section the tubes, instead of being circular, are crescentic or shield shaped. They taper in size toward the center, where they appear to be rather small, and are gently curved either upward or downward, it is impossible to tell which. None of the finer structures are preserved.

I am disposed to believe that this may be the same species as G. cylindrica var. robusta, differences in preservation causing these forms to present considerable difference in appearance. It is true that the typical example of G. cylindrica is hollow; yet, as other examples apparently representing the same species are solid, this difference probably would not hold. The mural tubes in G. cylindrica are normally, perhaps always, circular in cross section, yet the rhombic shape of those of the present example may be distorted by compression or by mutual crowding. This specimen is considerably larger than the type of G. cylindrica, but not so large as an example provisionally referred to the variety robusta, and it is not conceived that this difference would have much weight if an agreement were found to exist in other particulars.

A second specimen referred to this species is much smaller and with finer tubes.

Horizon and locality.—Middle of Capitan formation, peak north of Pine Spring, Guadalupe Mountains, Texas (station 2902).

GUADALUPIA? sp.

Pl. V, figs. 7 to 11.

In the highest fossiliferous horizon of the Capitan limestone occurs a form whose zoological affinities are obscure. Several sections are represented by figs. 7 to 10 of Pl. V. The original shape seems to have been short-cylindrical, open at one end and closed by a rounded wall at the other. It is true that most of the sections are nearly circular, from which a spherical shape would be inferred, but some are elliptical and several are elliptical with one end open. If the latter are not broken or misshapen my interpretation as to the real configuration would appear to be demanded.

These bodies are of appreciable size, with a diameter of 1-1/2 mm. or less. The walls are rather thick and pierced by large circular pores extending through to the inside. Somewhat depends, however, on the interpretation of the sections. Some of these appear as hollow rings and some as disks having a rather coarsely reticulate structure, the "pores" being round. Intermediate conditions are found, showing clearly that the reticulate sections merely represent more tangential views through the same organism. Nearly always there is an outer envelope of dark-colored material to be distinguished alike from the surrounding rock and the included transparent calcite, which represents, as I suppose, the original test. At this horizon the large Fusulinas, and doubtless the smaller organisms, are completely covered by a coating, more or less thick, composed probably of dolomite deposited about them before they were buried by the calcareous sediments. In sections which show a reticulate structure the filling of the openings is of the same dark (dolomitic) material of which the outer coating is composed. Where the section shows an annular structure there is usually an inner as well as an outer coating of dolomite. If these sections were exactly perpendicular to the axis they would doubtless show alternating radial bands of light and dark, representing walls and pores, respectively, but this has rarely been observed. Almost always, owing to obliquity of sections, the pores are represented merely by regular scallops, sometimes on one but usually on both sides of the ring-shaped section, which very seldom pierce quite through the transparent testaceous or probably pseudotestaceous material.

Just what place in the animal kingdom these bodies occupy is a matter of doubt. That they are not foraminiferal is clear. I doubt that they are radiolarian, not only because of their supposed shape but because of their large size. It is possible that they may prove to be something in the nature of calcareous algæ, but the most probable hypothesis seems to be that they are calcareous sponges related to Guadalupia or possibly belonging to that genus. Their very small size is unfavorable to such a hypothesis, as is also the absence, so far as known, of any spicular structure or of a spicular outer layer, as in Guadalupia.

The form and structure of these small bodies somewhat recalls Schwager's genus Margaritina. Several important differences are, however, at once noted, since the Guadalupian form is smaller, is not inflated, and appears to have one end rounded and closed, with the other open. Even were the resemblance still closer it would not resolve the doubts in which its zoological relations are involved, since the position of Margaritina is not known with certainty.

Fig. 11 of Pl. V represents a similar though much more delicately constructed organism, which appears to have had a conical shape. Whether this should be regarded as belonging to the same or a different type can not be told. The small size and correspondingly diminutive construction of this body render the probability much greater than in the other case that it may be a radiolarian.

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

GUADALUPIA? sp. var.

Pl. XXVII, figs. 11 and 12.

The form which I have referred to merely as Guadalupia? sp. occurs near the top of the Capitan limestone, where it is associated with abundant Fusulina elongata and with sponges. At another somewhat distant locality, belonging probably to a different horizon, and in a different faunal association, since Fusulina elongata does not occur there, is found a form which much resembles Guadalupia? sp. The fossils from the latter station (2964), while in very close agreement with the others, show certain intrinsic differences, of a minor character, it is true, but such as for the present indispose me to refer both without reserve to the same species. While individual specimens can be selected in which these differences are not found, yet as a rule the forms under consideration differ in having the walls thicker and the length greater. No section, for example, like that shown by fig. 11 of Pl. XXVII, was obtained from the Guadalupe Mountains, where, in fact, most of the sections were nearly circular. This circumstance is possibly to be accounted for through drifting of the specimens by current or wave action, as a result of which the Fusulinas, as already mentioned, occur with their axes more or less parallel.

This form, by reason of its, thicker walls and more pronounced cylindrical shape, resembles Margaritina still less, than the foregoing species.

Horizon and locality.—Delaware Mountain formation, southern Delaware Mountains, Texas (station 2964).

Family POLYSIPHONIDÆ n. fam.

While it is difficult in the case of the only species which is known of this genus and family to distinguish the specific characters from the generic, it is still more difficult to give a family diagnosis which may be expected to stand, and to discriminate the family characters from those which are generic and specific. Probably all types which could be referred to this family would have a conical or cylindrical shape, a thin outer wall, porous possibly but without ostia, and an internal structure consisting of tubular canals, some of which run lengthwise and some in a radial direction.

Genus POLYSIPHON n. gen.

The characters of the singular form on which this genus and family are based are so unique that it is difficult, in the case of the single species known, to distinguish the generic characters from those which are more purely specific. Nevertheless, I would briefly note here some of the structural features which would presumably be present with some modification in all species of the genus.

The shape and general appearance of this form remind one of a zaphrentoid coral, as the shape is conical, with a depression in the upper or larger end semblable to a calice. Internally the structures consist of tubular canals, one of which occupies a central position, the others being arranged regularly about it. In the typical species the peripheral tubes are five in number. These bifurcate and by this means connect with one another, and they also connect with the axial tube by radially directed tubular ducts, which are developed at the same level. The outer wall and those of the tubes are thin; the tubes and intervening spaces are hollow.

At first I was disposed to regard this as a singularly preserved coral, but it would indeed be a peculiar preservation which would metamorphose the typical structure of a zaphrentoid coral into that of the organism above described. We can probably eliminate the corals entirely from the list of possibilities.

There is another group which I must mention in this connection, though its relationship at first seems so remote that I almost hesitate to introduce it—the Echinodermata. Pentameral symmetry is rather persistently maintained by this class of organisms, and in a silicified fragment which I think must be regarded as the proximal end of an echinoid spine somewhat similar structures are shown. Aside from structure, however, the size and shape of the specimens which form the subject of the present description are such that it seems to me rather unlikely that they belonged to a crinoid or an echinoid, although it can not be said that such an origin is impossible.

On the whole, the calcareous sponges of the order Sycones can best be made to receive this form, whose position would probably lie in the vicinity of Cystothalamia and Guadalupia, but scarcely in the same family. A new family must be provided, which may be called the Polysiphonidæ.

It is hardly necessary to give in detail the family characters which distinguish this singular type. The family Polysiphonidæ is quite distinct from the other families of the Sycones, the nearest being doubtless the Cystothalamiidæ. From this the Polysiphonidæ differ in having a solid instead of a perforated outer wall, in being without ostia, and in having a definite and peculiar arrangement of the internal tubular structures, the tubes in Cystothalamia being more numerous, imperfect, chiefly radial in direction, and without and definite order of arrangement.

Type.—Polysiphon mirabilis.

POLYSIPHON MIRABILIS n. sp.

Pl. XVI, figs. 11 to 11b.

This species is based on two specimens, which may originally have belonged to the same individual. In general they look much like a zaphrentoid coral, and I temporarily placed them with the rugosæ, attributing to some peculiarity of preservation the internal structures, which are quite unlike those of the co&elig;lenterates. From the outside, however, they show differences from the corals, because the exterior is smooth, without annular growth lines, and more especially without the fine longitudinal ribs which are connected with the development of the septa. On the interior the structure is still more unlike that of the corals, but it is also so different from the normal sponge structure that my reference to the sponges is not made without misgivings. There is a thin conical bounding wall, and in the larger specimen a depression in the upper end corresponding to the calice. Within, however, instead of plates we find tubes. There is an axial tube, which is closed at the top and makes an elevation or boss in the center of what presumably should be considered the cloaca. Around this there are in the present specimen five other tubes, which are in part free and in part adnate to the outer wall. Toward the top these tubes bifurcate, and thus connect with one another. They also connect with the axial tube by large radial ducts, developed at the same level just below the point of bifurcation. In the present specimen these tubes, except the central one, are open around the edge of the cloaca, but it is impossible to tell whether this is the original condition or whether the rim of the cloaca has been broken off and with it the upper or terminating portion of the tubes. The walls are thin, without at present any perceptible traces of spicules. The tubes themselves and the intertubular spaces are empty.

These structures have been described as if they were perfectly regular, and they are in fact remarkably so, but slight deviations from the ideal scheme above described do occur, and they are represented in my figures.

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

Family CYSTOTHALAMIIDÆ n. fam.

The difficulties which were met with in the case of the Polysiphonidæ in defining the family characters when only one species is known occur again with the Cystothalamiidæ. Genera so related to Cystothalamia as to be referable to the same family would probably have a cylindrical or conical form, the upper end depressed into a more or less profound cloaca, which does not, however, persist to the bottom. The external wall is thin and perforated by pores of two sizes and by ostia. The internal cavity is occupied by cysts, which have a more or less irregular arrangement, and in especial are not grouped into separate rings, as in the Sphærosiphoniidæ and the Sphærocœliidæ. It is the absence of this character, as well as the presence of others, which distinguishes the three families of Guadalupian Sycones here discriminated and named from those mentioned above. They are so distinct from one another and from such other zoological families as I have found distinguished among the Sycones that it has hardly seemed necessary to indicate specifically the characters on account of which they are believed to demand separate recognition.

Genus CYSTOTHALAMIA n. gen.

This genus grows as rather slender, cylindrical, branching stems. It is probable, but not certain, that a more or less profound cloaca was present in the upper end. The internal structure consists of vesicles, which toward the periphery assume to a greater or less degree regularity of size and shape, while toward the interior they are more irregular, but they do not form regular superimposed chambers or tiers of chambers, as in the Cœlosiphoniidæ. Being built up by aggregations of vesicles, the exterior usually, perhaps normally, has a mammillate surface. The vesicles communicate with one another by means of circular openings and to some extent by tubelike canals. With the surrounding medium, they communicate by means of numerous small, regularly arranged openings in the outer walls and also probably by a series of larger and less numerous pores, or ostia, which sometimes project as little tubes.

The walls are thin and, at present, apparently structureless, but doubtless originally consisted of an entanglement of spicules.

I have felt compelled to propose a new family name for this genus, because among the imperfectly known Paleozoic Calcispongiæ I have been unable to find one described in which it could appropriately be located. It can not be placed in the Sphærosiphoniidæ or the Sphærocœliidæ, because it is not made up of a series of rings or segments, nor, in spite of a certain analogy, can it be placed with the Guadalupiidæ, because it is not composed of discrete, separate-walled tubes and because it lacks (?) an outer spicular or fibrous layer. The most significant features of the Cystothalamiidæ at present appear to be the absence of a persistant tubular axis, the continuous nonsegmented augmentation, and the camerate structure, consisting of apparently hollow cystiform chambers.

Type.—Cystothalamia nodulifera.

CYSTOTHALAMIA NODULIFERA n. sp.

Pl. VII, figs. 1 to 3; Pl. XXXI, figs. 20 to 20b.

This species grows in irregularly cylindrical, more or less contorted forms, having a diameter of 7 to 10 mm. or more. That these stemlike bodies are sometimes forked is indicated by several specimens. In a silicified example there is a central tube extending part way down the organism, which would seem to be a cloaca, although I am not sure that this is a constant feature. Many specimens do not show this structure, which, however, would be confined to the terminal portion of branches. Externally, the surface is covered with small rounded elevations or monticules, and in one specimen by spoutlike tubular projections. In this case the stem is large and probably old, and it lacks, over this part of the surface, the little monticules which smaller branches show. This appearance may be accounted for by supposing it to belong to old age, the branches having been thickened so as to obliterate the monticules and to prolong into short tubelike projections the ostia soon to be described.

The surface is covered with numerous small circular openings, rather regular in size and distribution, separated by intervals about two, but sometimes one or three times their own diameter. There are also other openings, which, as seen in a silicified specimen, may possibly be holes broken through the wall, but are, I judge, really ostia and proper features of the sponge. In old portions they appear to be extended into tubes, as above described.

The internal structure consists of cysts, more or less equal in size, regularly arranged around the periphery. They manifest no tendency to an arrangement in circular series, and wherever regularity is shown the series are more diagonal than transverse. In cross sections, so far as observed, there is no uniformity in their size, shape, or arrangement. In tangential sections, however, they tend to be rather regular in these particulars, having somewhat of a rhombic or, more exactly, a shield-shaped outline, especially near the surface. From this, however, they pass locally and probably regularly toward the middle into cavities of larger size and irregular shape. The walls in section are frequently incomplete, showing that the chambers communicate freely, and in the silicified specimen above referred to the chambers are seen to open into one another by rounded pores, while here and there a few tubes, whose direction is approximately axial, though more or less slanting, probably served as an additional means of circulation. It is the outer walls of these cysts which produce the mammillate surface of the sponge already noted.

While it may be supposed that this organism was composed of spicules, absolutely no trace of such structures remain so far as my observations go.

A first examination of this sponge conveys the impression that it is very widely different from the genus Guadalupia, while an axial view of the tangential section might mislead a casual observer into confusing them. The two sponges may, however, be really somewhat more nearly related than at first appears, while they are far from belonging to the same genus. The outer surface of Cystothalamia, with its monticules and little pores, as seen in the silicified specimen, certainly does not resemble the reticulate spiculous surface of Guadalupia. Specimens from the Capitan formation referred here show, in fact, no superficial structure at all, neither the spicules of one form nor the pores of the other. The internal structure also at first seems to possess as little resemblance, but if one considers the cysts as modified in shape so as to form tubes, or the tubes, in view of their being separated by diaphragms, as composed of several cysts arranged in linear series, a structure not far removed from Guadalupia would be produced. Of course there still remain some perhaps fundamental differences, such as the intercommunication between the chambers of Cystothalamia and the fact that the tubes in Guadalupia have complete and separate walls, while in Cystothalamia the chambers have only partial walls, the upper surface of one partition answering for the lower part of the next, or, at least, such appearing to be the case.

In spite of this very doubtful analogy between them, I am not including Cystothalamia and Guadalupia in the same family—that of the Guadalupiidæ—although it may be that such a course should be followed. While related to Amblysiphonella, Sollasia, etc., Cystothalamia differs from the Sphærosiphoniidæ in not being made up of regular superposed chambers or systematic annular groups of chambers.

Horizon and locality.—Top of Capitan formation, Capitan Peak, Guadalupe Mountains, Texas (stations 2905? and 2966). Delaware Mountain formation, southern Delaware Mountains, Texas (station 2964). Delaware Mountain formation, Comanche Canyon, Glass Mountains, Texas (station 3763).

CYSTOTHALAMIA? sp.

Pl. VII, fig. 5.

This species is represented by the single specimen illustrated on Pl. VII. It consists of a complex or more or less spherical bodies arranged about an axis so that they form an approximately straight stem. Their outward portions are regularly curved, but toward the axis they are deformed by mutual contact. The minute structure of this organism is no longer preserved. Portions of the surface seem to be covered with small rounded elevations or tubercles, which may be taken for ostia. Through the nongranular calcite which composes the bulk of the organism small opaque spots, probably of dolomite, are distributed, but it can not be told whether these are inorganic grains or represent fine tubular or porous structure. Similar appearances have been noted where it was difficult to tell whether the substance was of organic or inorganic nature.

In a preliminary account of the Guadalupian fauna I referred this organism to Mammillopora, or at least to the type which King includes under that name in his account of the Permian fossils of England. That it is not congeneric with M. mammillaris, however, I am now fairly assured, for it consists of an aggregate of several discrete spherical bodies, while Mammillopora appears to be a single homogeneous organism, growing, however, in a more or less mammillate shape. Moreover, Mammillopora consists of a solid spicular network, while the form under consideration was, I suspect, made up of hollow shells, the present fossil, which is solid, being a filling up of these chambers, and its apparent structure not really organic. It is only on the latter supposition that this form can be even provisionally placed with Cystothalamia. Should the interpretation adopted be correct, however, there might well be some doubt as to the propriety of placing this sponge in that genus. Its proper position may be in the Sphærocœliidæ, somewhere near Sollasia. In fact, it might with equal propriety be placed with that genus, but it is doubtful if it really belongs to either.

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

Family SPHÆROSIPHONIIDÆ Steinmann.

Genus AMBLYSIPHONELLA Steinmann.

AMBLYSIPHONELLA GUADALUPENSIS n. sp.

Pl. VII, figs. 7 to 8a,

This species grows in the usual cylindrical shapes, reaching a diameter of about 16 mm., and so far as known it occurs in single stalks and not in colonies. The cloaca has a diameter of 4 mm. The exterior is more or less marked by constrictions, but the superposed rings of which the organism is constructed do not, as such, show clearly upon the exterior, which appears to be irregular or verrucose. The interior of the structural rings is divided into simple cysts, which are large and not very numerous. The walls are thick, and at present appear dense and structureless, the outer and inner walls being perhaps a little heavier than those of the cysts.

But two specimens have come to hand, both of which are represented in my figures. That furnishing a natural longitudinal section may be taken as the type, as it best shows the structures characterizing the genus. The other specimen affords a better idea of the configuration, though possibly misleading in appearing to be verrucose on the exterior. From this example was cut the thin transverse section which is illustrated by my figure. It can not be determined that this specimen belongs to the same species as the type, but this is believed to be the case.

The only American species of Amblysiphonella known at this time is A. prosseri Clarke, from the Pennsylvanian of the Mississippi Valley. From this species A. guadalupensis is clearly distinct, by reason of its greater diameter, less strongly annulated shape, lower chambers, and thicker walls.

A. guadalupensis is a smaller form than any of the four species described by Waagen from the Salt Range, but it most nearly resembles that to which he gave the name A. multilamellosa. The much less numerous and much thicker walled cysts of the American form distinguish it. It differs from the other Indian species not only in smaller size, but in relatively smaller cloaca, more massive walls, and more equal vesicle and tube walls. In some respects it more closely resembles the typical species A. barroisi,a differing, however, in having a relatively smaller cloaca, thicker walls, less strongly annulated exterior, and more copious vesicles(?).

---

aNeues Jarbuch, 1882, vol. 2, p. 169, pl. 16, figs. 1-1d.

---

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

Family SPHÆROCŒLIIDÆ Steinmann.

Genus STEINMANNIA Waagen and Wentzel.

STEINMANNIA AMERICANA n. sp.

Pl. VII, figs. 4 and 4a.

Of this interesting form there has come to hand only one specimen. It has a subcylindrical shape, somewhat regularly interrupted by gentle constrictions, and is slightly curved. It is incomplete at both ends, the present length being 23 mm. The diameter of the larger end is 9 mm. and that of the smaller 7 mm. The constrictions correspond in a general way to internal partitions, which are about 3 mm. apart. There is no central cloaca, and the presence of a small osculum in the center of each partition has not been ascertained. The flattened segments made by the gently curved partitions are apparently not entirely hollow, but the definition of the thick walls and of whatever internal structures are present is not very distinct. There is certainly a much less extensive development of vesicular tissue than in the Indian specimens belonging to this genus.

The surface is perforated by small, round openings, while the structure of the test as a whole appears to be finely porous. Ostia are apparently absent. The larger pores, and probably the smaller ones also, appear to be a structural feature of the partitions as well as of the outer wall.

I have been in some uncertainty whether to place this species with Waagen's genus Steinmannia or with Steinmann's Sollasia; but the data at hand seem distinctly to favor a reference to the former. The structures in my specimen are not very clear, the test and the infilling whitish matrix being of about the same color and without any sharp boundaries between them. Such partitions as may be said to divide the otherwise hollow chambers have an appearance less of vesicles, as presented by my specimen, than of divisions more or less parallel to and near the upper or lower wall. Because of the probable presence of internal walls, although they are few, and the absence of ostia, it seemed best to place the Guadalupian species with Steinmannia. It is, however, strongly distinguished from the Salt Range species by the flattened or discoidal instead of spherical shape of the segments. The larger pores are of greater size in the American form, though it is really a smaller species and they are relatively much larger than the smaller pores.

From the exterior this form might be mistaken for a weathered example of Guadalupia cylindrica, but the tangential section shows the structure to be very different.

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

Genus SOLLASIA Steinmann.

SOLLASIA? sp.

Pl. VII, fig. 6.

This division includes three small spongoid bodies, one from station 2926 and two from station 2966, both in the white limestone of the Capitan. They have a linear, somewhat moniliform shape, but the swellings in the specimen shown by fig. 6 of Pl. VII are unusually elongate. In the other specimens the nodes are more nearly spherical. In both cases what seems to be the filling of an original, rather thin-walled capsule is structureless, and the walls themselves are represented by fibrous or granular dolomite. No spicular structure can be now made out, and the obscure structures which at present replace the original walls I prefer to regard as inorganic.

These small bodies in their general nodular or moniliform shape strongly suggest Steinmann's genus Sollasia, but they do not show any evidence of being provided with ostia, nor, furthermore, of being divided at the constrictions by partitions, thus failing to conform in one essential particular to the Sphærocœliidæ. They probably belong to the Calcispongiæ, but, as I have already indicated, are very doubtful representatives of Steinmann's genus.

Horizon and locality.—Top of Capitan formation (station 2966), middle of Capitan formation (station 2926), Capitan Peak, Guadalupe Mountains, Texas.