246 THE FLORA OF THE DAKOTA GROUP. 



one of wliicli, C. lohdo-cremta (loc. cit., PI. XLI, Fi<>-. 1), is closely related 

 to Cissites ingens Lesq., mentioned above from the Dakota Grou}), and also 

 to Vitis Bnmeri, described by Prof Ward in the Laramie floi-a, with three 

 other species of the same genus. From the Green River Grouj) Ave have 

 one Cissus and AmpcJopsis fertiaria closely allied to A. qnhiqaeJoVui of the 

 present North American flora. 



The reference to the family Hamamelidese of the leaves of the Dakota 

 Group described under the generic name of Hamamelites, seems legitimate, 

 for it has been indicated by the author of the genus, Saporta, for two species 

 first described as Ahvm and Alnites Lesq. (Cret. FL, PI. lxii) and later as 

 Hamamelites quercifolms and H. cordatus Lesq. (Cret. and Teit. Fl., p. 71). 

 And indeed the leaf of this last species figured (loc. cit., PI. iv, Fig. 3), 

 compared to some of those of the living HamameUs virginka Linn., shows a 

 striking affinity and the evidence at least of a family relationship. Five 

 species of Hamamelites are described from the Dakota Group, and two 

 from the Upper Cretaceous of Montana. One is recorded in the flora ot 

 Gelinden, and H. fofJiergiUoides Ha\). is described from the flora of Sezanne, 

 and has been identified in the flora of the Laramie Grou]). 



Leaves of Cornus are generally knoAAU by their peciiliar aerodrome 

 nervation. The species of tlie genus are about equally distributed in the 

 geologic periods of Europe and North America, beginning in the Dakota 

 Group, by C. jirrecox related to C. Forchammeri Heer, a species described by 

 that author from the schists of Atane, and also recognized later in the Upper 

 Cretaceous of Montana. The type is preserved in C. NidfaUii Audub., now 

 living in California, and C. asperifolia Michx., of the Atlantic slope of North 

 America. The flora of Patoot has two species, one of which is also found 

 among the specimens from Montana, evidently showing the affiliation of the 

 species in the Cretaceous stages with the original ty])e in the Dakota Group. 

 In more recent formations we have four species in the Laramie flora, one of 

 them also identified ij^ the Miocene and one in close relation to leaves of 

 the auriferous gi-avel deposits of California. From the Tertiary of Europe 

 twelve s])ecies are described, one only from Sezanne. Li the flora of our 

 epoch about twenty species are knoAvn to ])otanists; of these foixrteen lielong 

 to the North American flora, six pertaining exclusively to that of the Pacific 

 States. 



The family of the Magnoliaceae may be considered as the most inter- 

 esting of the ])aleontological series of jilauts. The genus Magnolia first, is 

 represented in the Dakota Group 1)\- a conical, cylindrical Itranch of fruit- 



