8fi FOSSIL FLORA OF THE JOHN DAY EASIN, OREGON. [bi'i.l. 201. 



th(> lony- lol)o :iik1 the (>i)po.sitc of the l»hid(>. The iiuijor part of the 

 l)l:ide is onl}" o em. broud. The .sniolo husal h)he is 2 em. lono-. 



This leaf presents a very eurious appearanee with the h)ng-, sharp- 

 pointed lobe on one side, and it is (luitc possible that it is an a])noi'mal 

 exampk^ perhaps of sonu^ well-known form. If it is the normal form, 

 it is eertainly one that will l)e readily reeognizable in future. Its 

 affinity is obviously in doul)t. If it were not for the presence of the 

 single basal lobe it mio-ht perhaps be referred to Betula, })ut I do not 

 know an}" speeies, either living' or fossil, with which it can be satisfac- 

 torily compared. 



Locality. — White hill one-half mile east of original Van Horn's 

 ranch locality, V2 miles west of Mount Vernon, Grant County, Ore- 

 gon. Collected by Knowlton and Merriam, July, 1901 (U. S. Nat. 

 Mus., No. 8537). 



Phyllites inexpectans n. sp. 



PI. XVI, fig. 6. 



Leaf evidentl}' coriaceous, obovate-lanceolate in shape, narrowed 

 below, obtuse and rounded at apex; margin perfectly entire; petiole 

 long, very strong for the size of the leaf; midril) also very strong; 

 other nervation obscure. 



This form is represented by the example figured. It is narrowly 

 obovate-lanceolate in outline, 3.25 cm. long and 9 mm. wide at the 

 broadest part, which is near the upper extremit3^ The petiole is 

 6 mm. long. Unfortunately nothing of the nervation except the thick 

 midrib is preserved. 



This leaf resembles quite closely certain of the leaves referred to 

 Quercus convexa Lesq. , '^ from the Auriferous gravels of California. 

 These leaves are uniformly thick and coriaceous, yet the nervation 

 shows distinctly in all cases. Our leaf is apparently no thicker, yet 

 has no trace of the nervation beyond the midrib. The petiole in Q. 

 convexa is uniformly short, whereas in the leaf under discussion it is 

 several times longer, being about one-fourth the length of the blade. 



On the whole, this leaf is perhaps closest to PhyJUtes ohscurus 

 Knowlton,^ from the Pa3^ette formation of Idaho. This latter spe- 

 cies, however, is elliptical-lanceolate or slightly ovate-lanceolate, while 

 ours is distinctly obovate-lanceolate. The petiole and midrib are simi- 

 lar, and both also lack details of nervation. Perhaps a larger series 

 might show them to grade together. 



Locality. — Van Horn's ranch, about 12 miles west of Mount Ver- 

 non, Grant County, Oregon. Collected by Knowlton and Merriam, 

 July, 1901 (U. S. Nat. Mus., No. 8547). 



"Foss. PI. Aurif. Gravel, p. 4, PI. I, figs. 13-17; Cret. and Tert. FL, p. 265, PI. XLV, figs. 5, 6. 

 6 Eighteenth Ann. Kept. U. S. Geol. Surv., Pt. Ill, p. 735, PI. XCIX, tig.s. 10, 11, 1898. 



