172 THE FLORA OF THE DAKOTA GROUP. 



The relation of tliesc leaves is remarkably well marked with those of 

 Rhanmus cohihrii/oidcs Ett. (Tert. Fl. v. Hjiriug, p. 7o, PI. xxv, Figs. 3-5), 

 better represented in Heer's Fl. Tert. Helv., vol. 3, j). 78, PI. cxxiii, Figs. 

 24-26, in leaves of" various sizes with a nervation of the same type. 



Habitat : Ellsworth County, Kansas. Nos. 437, 439, 534, 53(i, 796, 

 and 799 of the museum of the University of Kansas. Collected by E. P. 

 West. 



Order CELASTRINE^. 



Celastkophyllum decurbens, sp. uov. 

 PI. XXXVI, Fig. 1. 



Leaves subcoriaceous, large, lanceolate, gradually tapering upward 

 (point broken), narrowed and decurring along the petiole, serrulate, median 

 nerve strong; secondaries very close, oblique, craspedodrome. 



The leaf is 11™ long and 4*^"" l)road above the base, where it is enlarged, 

 minutely sen-ate, the borders being notched witli short equal teeth turned 

 upward, a mode of division which is very rare in leaves from the Dakota 

 Group. From below its broadest part it is narrowed and entire to the base 

 and decurrent, forming a narrow rim along the petiole, which is 2*^'" long. 

 In the upper part it ta])ers to the api)ai-ently acuminate apex. The second- 

 aries are very close and very thin, 2™'" distant, crossed by thin nervilles, 

 and forming a loose, square, or quadrangular areolation resembling tliat of 

 some Cenomanian species of Ficus, especially F. atar'uia Heer (Fl. Foss. 

 Ai-ct.,vol. 3, pt. 2, PI. XXX, Figs. 3, 3b; vol. 6, Abth. 2, p. 69). 



By its form, size, and serrate borders the species much resembles C. 

 lanceoldfiiiii Ett.,' a fragment, of which both the lower and upper part are 

 destroyed and which has the bt)rders more distantly serrate, the secondaries 

 twice as distant, though at the same angle of divergence of 40°, and the 

 same type of areolation; the median nerve, however, is much thicker. 

 Another fragment figured in Saporta and Marion' is referred to the same 

 species. It shows the teeth still larger than in Ettingshausen's iigure, a far 

 different areolation, and a thick, short jietiole slightly winged. That the 

 fragments from Gelinden may represent a different species from that from 

 Niederschoena, as the authors suppose, does not concern my present 



' Kieideflora v. Niorterschoena, p. 260, PI. iii, Fig. 9. 

 . ''Vffx. M.irnes Heera. Geliiidou, PI. XII, Figs. 1, 2; and Rdvis. Fl. Heers. Geliiiden, PI. xiv, Fig. 2; 

 C Bmedciii, Sap. & Mar. 



