Geological Survey Bulletin 204 Fossil Flora of the John Day Basin, Oregon

Family FAGACEÆ.

FAGUS? sp.

FAGUS CASTANEÆFOLIA Unger. Lesquereux, Proc. U. S. Nat. Mus., Vol. XI, p. 18, 1888.

This determination was based on a single fragment of the lower portion of a leaf. It is doubtful even whether it belongs to Fagus, and can have no value in fixing the age of or affording a stratigraphic mark for these beds.

Locality.—Van Horn's ranch, about 12 miles west of Mount Vernon, Grant County, Oregon. Collected by Maj. Charles E. Bendire (U. S. Nat. Mus., No. 2466).

QUERCUS FURCINERVIS AMERICANA Knowlton.

QUERCUS FURCINERVIS AMERICANA Knowlton, Cat. Cret. and Tert. Pl. N. A., p. 192, 1898.

Quercus furcinervis (Rossm.) Ung. Lesquereux, Proc. U. S. Nat. Mus., Vol. XI, p. 22, 1888.

This is not the European form, and is best indicated by the varietal name given above. This determination is based solely on a part of the middle portion of a single leaf, which is very similar indeed to Castanea pulchella Knowlton a from Lower Miocene beds in the Yellowstone National Park.

Locality.—Cherry Creek, Crook County, Oregon. Collected by Maj. Charles E. Bendire (U. S. Nat. Mus., No. 2554).

---

a Mon. U.S. Geol. Survey, Vol. XXXII, Pt. 11, p. 702; Pl. LXXXVII, fig. 1

---

QUERCUS? sp.

Pl. VIII, fig. 4.

Leaf very thick, narrowly obovate, obtusely wedge-shaped at base (apex destroyed); margin entire below, probably toothed above; midrib very thick, especially below, secondaries also strong, 6 or 7 pairs, alternate at an angle of about 45°, a pair near the middle of the blade largest, with strong outside branches, probably ending in lobes or teeth; nervilles very strong and deeply impressed, both broken and percurrent; finer venation producing large areas.

A single broken specimen is all that I now refer to this form. It was clearly a large thick leaf, with prominent deeply impressed nervation. It is entire in the lower portion, but from the size and disposition of the secondaries in the middle of the blade it was apparently lobed or toothed above. Its length was about 12 cm. and the width about 6.5 cm. It is doubtfully referred to Quercus.

Locality.—Cherry Creek, Crook County, Oregon. Collected by Knowlton and Merriam, July, 1901 (U. S. Nat. Mus., No. 8555).

QUERCUS PAUCIDENTATA Newb.

QUERCUS PAUCIDENTATA Newb., Proc. U. S. Nat. Mus., Vol. V, p. 505, 1883; Later Extinct Floras, p. 76, Pl. XLIII, fig. 1, 1898.

It is with some misgiving that this species is permitted to stand. It is based, as Newberry says, on a single example, the only one ever found. It is in all probability a large leaf of Q. affinis, but rather than complicate matters by combining forms without sufficient material I have preferred to keep it distinct.

Locality.—Bridge Creek, Grant County, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus., No. 7059).

QUERCUS DRYMEJA Unger.

QUERCUS DRYMEJA Unger. Lesquereux, Cret. and Tert. Fl., p. 245. Pl. LIV, fig. 4, 1883.

Lesquereux has figured a single leaf of this form, and the Bendire collection contains another example. Neither of these is perfectly preserved, but they seem to differ from other allied forms. It may be, however, that they are very large leaves of Q. consimilis Newb.

Locality.—Bridge Creek, Grant County, Oregon. Collected by Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 9231, 9328).

QUERCUS SIMPLEX Newb.

QUERCUS SIMPLEX Newb., Proc. U. S. Nat. Mus., Vol. V., p. 505, 1883; Later Extinct Floras, p. 78, Pl. XLIII, fig. 6, 1898.

As stated by Newberry the collections from Bridge Creek always contain a large number of leaves of this species. They are of the same size and shape as leaves of Q. consimilis, and only differ in being entire margined, and Newberry questioned as to whether they might not be only a varietal form of that species. As none of the specimens that have passed under my observation show any such gradation, it may be taken as established that they are distinct.

Locality.—Bridge Creek, Grant County, Oregon, collected by Rev. Thomas Condon (U. S. Nat. Mus., No. 7058a) and Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 9219, 9227, 9238, 9249, 9252, 9264, 9283, 9289, 9295, 9300, 9302, 9312, 9316, 9332, 9336, 9369, 9380, 9383). Officer's ranch, lower end of Butler Basin. Collected by Dr. John Merriam, July 22, 1901 (U. S. Nat. Mus., No. 9210).

QUERCUS AFFINIS (Newb.) n. comb.

Fraxinus affinis Newb., Proc. U. S. Nat. Mus., Vol. V, p. 510, 1883; Later Extinct Floras, p. 127, Vol. XLIX, fig. 5, 1898 (1899).

Quercus furcinis Rossm. Lesquereux, Cret. and Tert. Fl., p. 244, Pl. LIII, figs. 10— 12, 1883.

Quercus Breweri Lesq., idem, p. 246, Pl. LIV, fig. 9 (not other figures described as Q. Breiceri) 1883.

This species came first into the hands of Dr. Newberry and was called by him Fraxinus affinis. He appears to have noticed only a single example, namely, the one made the type, yet there is a larger leaf on the same piece of matrix. A little later a number of similar leaves from the same locality were studied by Lesquereux and by him identified with Quercus furcinervis of Rossmassler, and his own Q. Breweri (See synonymy above).

As Newberry well states, there is a strong resemblance between his leaf and the living Fraxinus americana, yet in placing it by the side of the figures given by Lesquereux there can be no doubt that only one species is represented, and moreover that this is much more like Quercus than Fraxinus. This view is further strengthened by the finding of additional leaves in later collections. It is therefore clear to my mind that they are oak leaves, and I have placed them under Quercus.

It now remains to explain the selection of the specific name for these leaves. Newberry's Fraxinus affinis was published March 21, 1883, whereas Lesquereux's volume, as pointed out on page 12, could not have been issued until late in 1883 or more probably not until some time in 1884. Newberry, therefore, clearly has priority. Now, if either of Lesquereux's references of these leaves to species of Quercus was valid, it would simply be necessary to transfer Fraxinus affinis to the one selected, but in my opinion they are not. I do not think that the leaves referred to Quercus furcinervis a are the same as Rossmassler's species, and, obviously, the leaf called Quercus Breweri b is not the same as the others with which it is associated, and is identical with the one on the preceding plate. On these grounds I have retained Newberry's specific name because it is the oldest, and I have placed them under Quercus because I consider them to be oak leaves.

Locality.—Bridge Creek, Grant County, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus. No. 7125). Type of fig. 9, op. cit., is in Mus. Univ. Cal., No. 1774.

---

a Lesquereux, Cret. and Tert. Fl., Pl. LIII, figs. 10-12.

b Idem, Pl. LIV, fig. 9

---

QUERCUS CONSIMILIS Newb.

QUERCUS CONSIMILIS Newb., Proc. U. S. Nat. Mus., Vol. V, p. 505, 1883; Later Extinct Floras, p. 71, Pl. XLIII, figs. 2—5, 7—10, 1898.

This species is very abundant in the collections from Bridge Creek, and is well characterized and figured by Newberry. It is closely allied to Q. Breweri Lesq., but can be readily distinguished by being, shorter and relatively much broader.

A single broken leaf was found by Merriam at Officer's ranch, in the lower end of Butler Basin.

Locality.—Bridge Creek, Grant County, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus., Nos. 7048, 7050, 7057, 7063, 7070) and Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 9230, 9235, 9239, 9247, 9251, 9259, 9261, 9271, 9276, 9280, 9282, 9287, 9305, 9311, 9321, 9329, 9333, 9337, 9342, 9344, 9347, 9351, 9362). Officer's ranch, lower end of Butler Basin. Collected by Merriam, July 22, 1901 (U. S. Nat. Mus., No. 9209).

QUERCUS BREWERI Lesq.

QUERCUS BREWERI Lesq., Cret. and Tert. Fl., p. 246, Pl. LIV, figs. 5-8 (non fig. 9), 1883.

Some years ago, when preparing the manuscript for my Catalogue of the Cretaceous and Tertiary Plants of North America, I was led from a casual examination of the figures to refer this species to Q. consimilis Newb. and Q. paucidentata Newb. Since that time I have found the long narrow leaves of Q. Breweri in the vicinity of Ashland, Oregon, and I have also gone carefully over the abundant material from Bridge Creek, which contains a number of finely preserved leaves, and I am now inclined to regard them as distinct from Q. consimilis. They are undoubtedly close to this species, yet differ in being much longer and narrower. They can be readily distinguished. I have seen and examined the types of this species in the University of California, and I am the more convinced that it is distinct.

Locality.—Bridge Creek, Grant County, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus., No. 9345), by Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 9218, 9224, 9277, 9294, 9322, 9331, 9345, 9352), and by F. H. Knowlton. Types in Mus. Univ. Cal., Nos. 1761, 1762, 1763.

QUERCUS PSEUDO-ALNUS Ettingshausen.

QUERCUS PSEUDO-ALNUS Ettingshausen. Lesquereux, Cret. and Tert. Fl., p. 244, Pl. LIII, figs. 1—7, 1883.

Populus polymorpha Newb., in part. Later Extinct Floras, p. 50, Pl. XLIX, fig. 7, 1898.

Under the above name Lesquereux has figured a number of leaves from Bridge Creek. They differ considerably among themselves, yet may well belong to a single polymorphous species. Nearly all the forms are abundant in any collection from this locality.

The leaf figured by Newberry as Populus polymorpha (loc. cit., fig. 4) is certainly the same as fig. 6 of Pl. LIII in the Cretaceous and Tertiary Flora, and I have placed it under this species as determined by Lesquereux. It is one of the most abundant forms in all collections.

Locality.—Bridge Creek, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus., No. 7051) and Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 9229, 9250, 9253, 9266, 9274, 9292, 9307, 9309, 9317, 9319, 9326, 9334, 9343, 9355, 9358, 9365, 9368, 9372, 9374). Types of original American figured specimens in Mus. Univ. Cal., fig. 1, No. 1767; fig. 2, No. 1768; fig. 3, No. 1769; fig. 4, No. 1770; fig. 5, No. 1771; fig. 6, No. 1772; fig. 7, No. 1773. Officer's ranch, lower end of Butler Basin. Collected by Dr. John C. Merriam, July 22, 1901 (U. S. Nat. Mus., No. 9204).

QUERCUS OREGONIANA, n. sp.

Pl. VI, figs. 2, 3; Pl. VII, fig. 1.

QUERCUS n. sp., Knowlton in Merriam, Univ. Cal., Bull. Dept. Geol., Vol. II, No. 9, p. 288, 1901.

Populus polymorpha Newb., in part. Proc. U. S. Nat. Mus., Vol. V, p. 506; Later Extinct Floras, p. 50, Pl. XLVI, fig. 4 (not the other figures of P. polymorpha).

Leaf membranaceous in texture, elliptical-ovate or slightly elliptical obovate in shape, rather abruptly rounded at base, obtusely acuminate at apex; margin coarsely and irregularly toothed, the teeth rounded or somewhat acute (petiole not preserved); midrib rather slender, perfectly straight; secondaries 8 or 9 pairs, alternate, arising at an angle of 45° or 50°, nearly straight, ending in the larger marginal teeth, often with one or two branches on the lower side, which also pass to the marginal teeth; the secondaries in the lower part of the blade closer than those above and at a less angle; nervilles numerous, per-current or broken, approximately at right angles to the secondaries; finer nervation producing a close, irregularly quadrangular network.

This species is based on two examples, one of which (Pl. VI, fig. 2) is the original of one of Newberry's types of Populus polymorpha, and the other (Pl. VII, fig. 1) a smaller specimen that was identified by Lesquereux as Carpinus betuloides Unger. The leaves are about 6 cm. in length and 4 cm. in width. The first-mentioned example is nearly perfect, lacking only the petiole, while the other lacks all of the basal portion. The drawing given in Newberry's Later Extinct Floras is not quite correct as regards the form, the teeth being uniformly rounded, and the nervation is only partially shown. The other example has never before been figured.

I am moved to take this leaf from Newberry's polymorphic aggregation for two reasons: First, because it differs from the other forms included under P. polymorpha; and, second, because I can not believe that it belongs with the genus Populus. It has much more the appearance and general facies of a Quercus, and for these reasons has been removed.

Locality.—Bridge Creek, Oregon. Collected by Rev. Thomas Condon (U. S. Nat. Mus., No. 7049) and Maj. Charles E. Bendire (U. S. Nat. Mus., No. 8484).

QUERCUS PSEUDO-LYRATA Lesq.

QUERCUS PSEUDO-LYRATA Lesq., Foss. Pl. Aurif. Gravel, p. 8, Pl. II, figs. 1, 2, 1878; Proc. U. S. Nat. Mus. Vol. XI, p. 17, Pl. X, fig. 1, 1888; Knowlton, Univ. Cal., Bull. Dept. Geol., Vol. II, No. 9, p. 308, 1901.

Quercus pseudo-lyrata acutiloba Lesq., Proc. U. S. Nat. Mus., Vol. XI, p. 17, Pl. XI, fig. 1, 1878.

Quercus pseudo-lyrata brevifolia Lesq., idem, p. 18, Pl. X, fig. 2.

Quercus pseudo-lyrata latifolia Lesq., idem, p. 18, Pl. XII, fig. 1.

Quercus pseudo-lyrata obtusiloba Lesq., idem, p. 18, Pl. X, fig. 3.

This species was originally described and figured by Lesquereux in his Fossil Plants of the Auriferous Gravel (p. 8, Pl. II, figs. 1, 2). After the description he has the following to say regarding the locality:

The locality is unknown, or at least not marked in the catalogue of the labels. The matrix of the specimens is a white, soft clay like that of the Chalk Bluffs of Nevada County, California, and no other species is preserved upon them except a fragment of a leaf apparently referable to Castanea intermedia Lesq. These specimens are evidently from the same formation and age as those of the Chalk Bluffs.

Both of the type specimens on which Lesquereux based his description and the above statement are preserved in the Paleontological Collection of the University of California (Nos. 1796 and 1796a), where I recently had the opportunity of examining them. A glance at the matrix was sufficient to show that they came from Van Horn's ranch, John Day Valley, Oregon. They form a part of the original Voy collection, made about thirty years ago, which fact is recorded on the back of each specimen. The matrix, mistaken by Lesquereux for a white, soft clay, is made up of very fine spicules of glass of volcanic origin, and is unmistakably that of the Van Horn's ranch locality. From this it appears that Quercus pseudo-lyrata was not originally, and, so far as now known, has never been found in California, or indeed outside of the John Day Basin. These facts are of great importance, since this characteristic species was relied upon to establish the correlation between the Auriferous gravels and the Van Horn's ranch deposits.

I have before nine all of the type and duplicate material, belonging to the United States National Museum, on which Lesquereux based the above enumerated varieties of this species, as well as the material obtained by Dr. John C. Merriam in 1900 for the University of California, and by myself in 1901. This comprises fully one hundred more or less perfect examples. There are, it is true, slight differences between the various forms, but I am now quite convinced that they are only individual variations, such as may be noted in the leaves of many species of living oaks. I have placed all these leaves in a single series and have found it quite impossible to draw any satisfactory line between them. They have consequently been referred to the single original form.

Locality.—Van Horn's ranch and vicinity, on South Fork of John Day River, 12 miles west of Mount Vernon, Grant County, Oregon. Collected originally by C. D. Voy (Univ. Cal., Nos. 1796, 1796a). Since collected by Maj. Charles E. Bendire (U. S. Nat. Mus., Nos. 2565, 2566, 2568, 2569, 2570), Dr. John C. Merriam in 1900 (Mus. Univ. Cal., Nos. 838, 839, 841, 842, 843, 844, 847a, 849, 852, 859, 865, 872), and by F. H. Knowlton, in July, 1901. (U. S. Nat. Mus., Nos. 8999—9015).

QUERCUS MERRIAMI n. sp.

Pl. VI, figs. 6, 7; Pl. VII, figs. 4, 5.

QUERCUS n. sp., Knowlton in MERRIAM, Univ. Cal. Bull. Dept. Geol., Vol. II, No. 9. p. 308, 1901.

Quercus pseudo-lyrata angustiloba Lesq., Proc. U. S. Nat. Mus., Vol. XI, p. 17, Pl. XI, fig. 2, 1888.

Leaves coriaceous in texture, narrowly lanceolate in outline, wedge-shaped at base, slenderly acuminate at apex; provided with 4 or 5 pairs of alternate or subopposite lobes, which are deltoid or deltoid-lanceolate in shape, usually sharp-pointed, but occasionally with the basal ones obtuse and rounded; petiole very long and slender; midrib moderately strong; secondaries usually at an acute angle, as many as the lobes and ending in their apices; intermediate secondaries few, apparently craspedodrome; finer nervation not well retained.

This species is now represented by more than twenty-five more or less perfect examples. The one shown in Pl. VII, fig. 4 was figured by Lesquereux under the name of Quercus pseudo-lyrata angustiloba. a As shown in Lesquereux's figure, it appears to lack the basal portion with the petiole, but this was covered by matrix, which has now been removed, exposing the long, slender petiole. The other specimens figured, as well as all but one or two of those now known, were obtained in 1901.

---

a This verietal name can not be remained, as it is preoccupied by Quercus angustiloba Al. Br., in Ludgw. Palaeontogr., vol. VIII, p. 103, Pl. XXXVI, fig. 3, 1861.

---

The first-mentioned example (Pl. VI, fig. 6) is about 14 cm. long, including the petiole, which is fully 4 cm. long. At the widest point between the lobes it is only a little over 3 cm., while at the narrowest point, which is near the middle of the blade, it is considerably less than 1 cm. The still larger example, shown in fig. 6, must have been 15 or 16 cm. long and 6 cm. broad between the points of the lobes. At the narrowest point it is about 2 cm. One of the smallest leaves is shown in Pl. VI, fig. 6. It is 9.5 cm. long, including the petiole of about 1.5 cm. in length. The broadest portion between the lobes is 2 cm., and the narrowest only 7 mm. in width.

The specimens representing this species are intimately associated with the numerous leaves of typical Q. pseudo-lyrata, and it is hardly to be wondered that Lesquereux, with only a single example before him, should regard it as an extremely narrow form of that species. But with the fine series now at hand it is clear that it is very distinct.

Among living species this form is certainly suggestive of Quercus heterophylla Michx. f.,the so-called Bartram oak, which is supposed to be a hybrid between Q. phellos and Q. rubra. This resemblance may be only superficial, but it is nevertheless plain.

I take pleasure in naming this species in honor of Dr. John C. Merriam, of the University of California.

Locality.—Van Horn's ranch and the two other near-by localities, 12 miles west of Mount Vernon, Grant County, Oregon. Type of fig. 4 collected by Major Bendire (U. S. Nat. Mus., No. 8505). Types of other figures collected by Knowlton and Merriam, July, 1901 (U. S. Nat. Mus., Nos. 8506, 8507). Type of fig. 7 collected by Merriam's expedition of 1900 (Mus. Univ. Cal., No. 846).

QUERCUS DURIUSCULA n. sp.

Pl. VIII, fig. 2.

Leaf coriaceous in texture, broadly obovate in general outline, deeply lyrate-pinnatifid into about five lobes, of which the basal are very small, triangular, and obtuse, the other broad, rounded, with deeply undulate or toothed lobes; midrib strong; secondaries three pairs, alternate, ending in the principal lobes, the upper ones with strong branches passing to the smaller lobes; finer nervation not well retained.

Unfortunately only one example of this form was found, and even this lacks a small portion of the base and has the upper lobes somewhat injured, evidently before fossilization. The length was about 5.5 cm. and the greatest width about 5 cm. The two basal lobes are less than 1 cm. in length. The outline and such details of nervation as are preserved are well shown in the figure.

This leaf clearly belongs to the white-oak group, and apparently finds its greatest affinity with Quercus minor (Marsh.) Sargent, the well-known post, or iron, oak, a species now common over much of the region east of the Rocky Mountains south of Massachusetts. It is so close to this species, in fact, that it can hardly be distinguished from many of the smaller leaves. It would therefore seem beyond question that the living Q. minor is a direct descendant of this fossil form, if, indeed, it has not remained practically unchanged from the time the Mascall beds were laid down to the present day.

Locality.—White hill one-half mile east of original Van Horn's ranch locality, 12 miles west of Mount Vernon, Grant County, Oregon. Collected by Knowlton and Merriam, July, 1901 (U. S. Nat. Mus., No. 8508).

QUERCUS URSINA n. sp.

Pl. VII, figs. 2, 3.

QUERCUS n. sp., Knowlton in Merriam, Univ. Cal., Bull. Dept. Geol., Vol. II, No. 9, p. 308, 1901.

Leaves coriaceous in texture, roughly obovate in general outline, 5 to 7 lobed, the lobes triangular or triangular-ovate in shape, very acute and apparently bristle-tipped; petiole long, relatively strong; nervation consisting of a strong midrib and as many alternate, rather thin secondaries as there are lobes; finer nervation not fully preserved.

This species is represented by several very well-preserved leaves. They are small leaves, about 4.5 cm. long, exclusive of the petiole, and about 4.5 cm. broad. The petiole is fully 1.5 cm. in length. The lobes, usually about 6 in number, are mainly triangular in shape, and are entire or occasionally with a single small sharp tooth, as shown in Pl. VII, fig. 2.

This species, so far as I am able to determine, finds its closest affinity among living species with Quercus nana (Marsh.) Sargent, the bear or scrub oak of the Eastern United States. The fossil form has a longer petiole, but otherwise the shape is very similar to certain of the smaller leaves of this species.

It is barely possible that these leaves may be only small forms of the polymorphous Quercus pseudo-lyrata, but I do not at present think so, for out of more than a hundred examples of the latter species there are no forms that can well be regarded as intermediate.

Locality.—Van Horn's ranch, about 12 miles west of Mount Vernon, Grant County, Oregon. Collected by Merriam's expedition of 1900 (type of fig. 2, in Mus. Univ. Cal., No. 841). White hill one-half mile east of above-mentioned locality. Collected by Knowlton and Merriam, July, 1901 (type of fig. 3, U. S. Nat. Mus., No. 8509).

QUERCUS DAYANA n. sp.

Pl. VI, fig. 1.

Leaf coriaceous, broadly elliptical-lanceolate, about equally obtusely acuminate at both base and apex; margin undulate, perhaps reflexed, otherwise entire; petiole very short and thick; midrib also very thick; secondaries thin and obscure, apparently about 12 pairs, alternate, emerging at a low angle, apparently ending in or very near the margin; finer nervation not retained.

The leaf figured—the only one thus far found—is 3.5 cm. long and 9 mm. broad and has the petiole only 3 mm. long. The outline and scant nervation are well shown in the figure.

This little leaf has such a familiar appearance that it would seem to be a known species, and, indeed, it does resemble more or less closely a number of forms, but after careful comparison I am forced to regard it as hitherto undescribed. It is clearly an oak leaf of the well known sempervirens type, and is allied to a number of fossil forms of this kind. It differs, for instance, from Quercus simplex Newb., a in being much shorter, relatively broader, with shorter petiole and closer, lower angled secondaries. It somewhat resembles Q. convexa Lesq., b of the Auriferous gravels of California, but differs in nervation, petiole, and other details. It is not greatly unlike the smallest leaves of Q. simulata Knowlton,c from the Payette formation of Idaho, but is sharper at both ends and has an undulate margin and a much shorter, thicker petiole.

Locality.—White hill one-half mile east of original Van Horn's ranch locality. Collected by Knowlton and Merriam, July, 1901 (U. S. Nat. Mus., No. 8546).

---

a Later Extinct Floras, p. 78, Pl. XLIII, fig. 6.

b Foss. Pl. Aurif. Gravel, p.4. Pl. I, figs. 13—17, 1878.

c Eighteenth Ann. Rept. U. S. Geol. Suv., Pt. III, p. 728, Pl. CI, fig. 3.

---

QUERCUS HORNIANA Lesq.

Pl. VIII, fig. 1.

QUERCUS HORNIANA Lesq., Proc. U. S. Nat. Mus., Vol. XI, p. 17, 1888 (the figure given on Pl. V, fig. 6, is not a figure of this specimen, being a figure of a specimen of Ulmus californica).

Castanea atavia Unger. Lesquereux, Cret. and Tert. Fl., p. 247, Pl. LII, fig. 2, 1884.

The type specimen of Lesquereux's Quercus horniana has not previously been figured, the figure supposed to represent it being that of a specimen of Ulmus californica from the same beds. It is here figured for the first time, and it needs but a glance to show that it is identical with the leaf identified by Lesquereux as Castanea atavia Unger, a fact apparently overlooked by Lesquereux when he established Q. horniana. I have not been able to see any European material of Castanea atavia, but a study of the type figures,d as well as others, leads me to the conclusion that Lesquereux was in error in identifying the John Day leaf with this species. In the European species the teeth are smaller and lower, the secondaries opposite and only about ten pairs instead of fifteen or more pairs, and the finer nervation is much more open. In the John Day form the margin is entire for a considerable distance above the base while the upper portion is provided with very large, sharp teeth.

While there can be no doubt that this leaf is the same as that figured by Lesquereux as Castanea atavia, there may be some as to its being referred to Quercus. The size and shape of the leaf, however, are very suggestive of an oak, and for the present it may remain in this genus. It is very well marked and one not likely to be easily mistaken for anything thus far discovered in these beds.

Locality.—Van Horn's ranch, South Fork of John Day River, about 12 miles west of Mount Vernon, Grant County, Oregon. Collected by Maj. Charles E. Bendire (U. S. Nat. Mus., No. 8504).

---

d Foss. Fl. v. Sotzka, p. 34 (164), Pl. X (XXXI), figs. 5-7, 1850.

---

QUERCUS ? sp. Knowlton.

Pl. VIII, fig. 3.

QUERCUS? sp., Knowlton in Merriam, Univ. Cal., Bull. Dept. Geol., Vol. II, No. 9, p. 308, 1901.

The collection made by the University of California contains a single leaf—the one figured—which appears to belong to Quercus. Only the basal portion is preserved together with a very short, thick petiole. The texture seems to have been coriaceous; the shape is lanceolate, wedge-shaped at base, and entire margined. The midrib is very thick and the secondaries—several pairs—subopposite. None of the finer nervation is retained.

This may possibly be a fragment of Quercus simplex Newb., which is so abundant at Bridge Creek, but it is only a fragment and the nervation is not well preserved, so I have hesitated to so regard it.

Locality.—Van Horn's ranch, about 12 miles west of Mount Vernon, Grant County, Oregon. Collected by Merriam's expedition of 1900. (Mus. Univ. Cal., No. 860.)