ANALYSIS OF THE DAKOTA GROUP FLORA, 251 



North America, three in the Western States, with one species of Negundo 

 each in tlie Western and Eastern States. Of tlie genns Negundo one 

 Miocene species is described from Em'ope and one from North America, 

 the latter from the Fort Union Group by Dr. Newberry. 



I have also described as Ncgtmdoides acntifolia in Cret. Fl. (p. 97, PI. 

 XXI, Fig. 5) fragments of two leaflets found in the Dakota Group, and 

 aj)parently part of a compound leaf, which I could relate only to the leaves 

 of Najando GaUfornirum Torr. and Gray. On account of the insufficient 

 characters of these fragments it is not possible to consider them as original 

 representatives of Negundo, nor even perhaps to admit them in the Aceracea;. 



The Sapindaceae have distinct representatives in numerous leaves of 

 Sapindus in the Cenomanian of Greenland and of North America. From 

 the schists of Atane Hear records two species of Sapindus, one of them, first 

 described from specimens of the Dakota Group, is also found in the Upper 

 Cx'etaceous of Patoot. No other species is known from the Ujjper Creta- 

 ceous, but the genus is represented in North American Tertiary bv eleven 

 species, two of them recorded from the Green River Group and two from 

 the Fort Union Group. At this epoch the Sapindaceae are mostly distributed 

 in the tropical regions. One species of Sapindus only remains in the North 

 American flora inhabiting the coast of Florida and Georgia. 



A numl)er of coriaceous leaves, whose relation to Celastrus is indicated 

 by their form and nervation, have been refei'red to this type under the 

 name of Celastrophyllum. Their characters are not clearly defined and 

 therefore their relation to living species of Celastrus is not positive. In the 

 leaves of the Dakota Group I have recognized three species of Celastro- 

 phyllum, one of which, C. ilecnnrns, is closely related to C. lai/reolafion Ett., 

 of the Cenomanian of Niederschoena, Saxony, and is also re})roduced with 

 distinct affinity in C. Benedeni Sap. & Mar., of the Senonian of France. 

 Heer has described tlu'ee species of this same generic division and one of 

 Celastrus from the Senonian of Patoot. The genus Celastrus is abun- 

 dantly rejH'esented in the more recent geological formations. From the 

 Laramie Group Prof Ward describes seven species of Celastrus, and from 

 localities referable to the same group I have recorded in Tert. Flora two 

 species of Celastrinites with three species of Celastrus, one Celastrinites 

 from the Green River Group and one Celastrus from the Miocene of Alaska. 

 Celastrus scandens Linn., is the only li\ang species remaining in the North 

 American flora. Though more than sixty species of Celasti-us, including 



