DESCIUPTJON OF SPECIES— PO]\lA(JE^. 297 



a number of living species from Cuba. Some of iliese, especially of llie family 

 of the Euphorbiacf'cB, Tricera retusa, Gray, T. fu&ciculosa, Gris., have a nerva- 

 tion and a texture of leaves exactly corresponding with those of the specimens 

 of Green River, and 1 now should be disposed to .rather refer them to ttiis 

 genus, or at least to the Euphorhiacece, abundant in the subtropical North 

 American flora, than to Australian types; for this Eucalyptus would be, like 

 the former, an anomaly in the Upper Tertiary flora of the Lignitic. As in 

 species of Tricera, the leaves are very sliort-petioled, attached to the stems 

 merely, as far as can be seen from the specimen, by the enlarged base of the 

 flat, broad midrib; the lateral veins, at an angle of divergence of 40°, pass 

 straight to the borders, where the}' join, with scarcely any curve, a distinct 

 marginal nerve, somewhat thinner than the veins. This apparent marginal 

 nerve is of course Ibrnied by the abrupt curve of the lateral nerves which 

 follow the borders, as more distinctly marked in fig. 11. In Tricera retusa, 

 we see exactly the same character, which is observable also in the distri- 

 l)ution of the numerous parallel secondary nerves, separated by thinner and 

 shorter tertiary veins, joined either in right angle by nervilles or in very 

 acute angle Ijy branchlets coming out from the midrib or from the lateral 

 nerves. From the fragments figured here, the leaves seem to be compara- 

 tively very long, for fig. 11 is twelve centimeters long and fifteen millimeters 

 broad; and, by comparison, the fragment represented in fig. 12, which is 

 more than one-third broader, should be part of a leaf about eighteen 

 centimeters long. 



Habitat. — Green River group, Wyoming, above fish-beds (Dr. F. V. 

 Hayden). 



ROSlFLORiE. 

 POMACES. 



Of the presence of plants of this family in our Tertiary flora, we have 

 as yet no positive evidence, the fragments described here as Cratcegtis being 

 too incomplete for positive identification. 



CRAT.S;GUS, Linn.' 



CrntiBg'iis? aequidcntatn, sp. nov. 



Plate LVIII, Figs. 4,4 a. 

 Leaves of large size, broadly lanceolate ; borders dentate ; nervation craspeilodromo. 



These fragments have the nervation of leaves of Crataegus, tlie lateral 

 nerves about equidistant and parallel, the lower branching outside, and the 



