ANALYSIS OF THE DAKOTA GHOITP FLOKA. 237 



Buttes, Wyoming. The Hora of the Green lvi\('i- Clroui) lia.s five species; 

 that of" the Miocene, eight; tour of them being also recorded from Ahiska, 

 the others from Oregon and ('ahfornia. From this it is observed that the 

 original type, 6". protecefoUa, is positively recognized in the Dakota Group, 

 but that its affiliation with more recent floras is not discovered until the Mio- 

 cene period is reitched, where its relation is marked with S. tenera Al. Br., 

 and later Avith a large number of the living- species. 



As remarked already, the first traces of dictyledonous leaves in the 

 flora of the world were discovered by Heer in the Lower Cretaceous (Neo- 

 comian) of Kome, Greenland, in fragments of leaves of the so-called Popidm 

 primceva Heer. One of the lea\es is preserved nearly entire, and upon the 

 same specimen there is an involucral scale which, thoug-h entire and without 

 hairs or cilia, apparently belongs, with the leaves, to a species of Populus. 

 The reference, which is generally admitted, is confirmed by tlie number of 

 leaves of species of Populus found in the Cenomanian. From the schists of 

 Atane, Heer has described four species, which are also recognized in the Da- 

 kota Group with four others, three of which are described by Dr. Newberry. 

 Three more are described as Populites from their analogy to leaves of Pop- 

 ulus, and belong also to the Dakota Grouj). As far as they are noAv known 

 the fossil species of Populus have been grouped in two sections according 

 to the type of nervation of their leaves: First, leaves with two pairs of 

 basilar primaries, the inner stronger, very oblique, curving inward in ascend- 

 ing; the outer or lower, shorter, generally parallel to the borders, with lowest 

 secondaries at a great distance above the primaries, not parallel to them; 

 second, leaves with lateral primary nerves open a nd lower secondaries about 

 equidistant and parallel with the primaries and upper secondaries. 



The first type is represented in the Dakota Group essentially l)y /'. 

 elliptka Newb. (Illustr. Cret. Tert. PI., PI. iii. Figs. 1, 2), whose leaves are so 

 remarkably similar in character to those of P. arefica Heer that this last 

 species, which is extremely common and variable, being mostlv iMiocene, 

 seems really a mere variety of the formei'. To the second tvpe are refer- 

 able the other species of Populus of the Cenomanian, mentioned aV)ove, 

 witli two species from Patoot, one of which is identified in the floi"a of 

 Atane and in that of the Dakota Group. In the Senonian of AVyoming two 

 species have been found, andfivein the upper Cretaceous of Vancouver Island. 

 From the Laramie Group as many as twenty-three species are recorded, 

 among them ten new ones described by Prof Ward in his Laramie Flora. 

 Of the whole lot twenty are of the first type or section, which may be called 



