ANALYSIS OP THE DAKOTA GKOUP FLORA. 247 



bearing carpels, and bj^ numerons well-preserved leaves, which are easily 

 ideutilied by size, I'orni, and nervation. Nine species are described from 

 leaves of this genus in the Dakota Group and four in the schists of Atane. 

 Of these, two pertain to both localities and two are recognized by Daw- 

 son in the Peace and Pine River series of Canada. None have been 

 ol)served in the Senonian of Patoot. But one of the species of the Dakota 

 Group, M. pscmh-aciimiuata, has been identified in the flora of the Upper 

 Cretaceous of Montana (Princeton collection). From the Laramie Group 

 five species have also been described and tAvo from the Upper . Miocene 

 beds of the auriferous gravel deposits of California. As the essential 

 chai-acters of the leaves of these difl'erent species are closeh" allied, .and 

 may be recognized in the successive tormations from the Cenoraanian to tlie 

 present epoch, the affiliation by gradual transition of different characters of 

 the species is put in full evidence. 



Some of the leaves of the Dakota Group, for example, are so remark- 

 ably similar to those of species of Magnolia of the present North American 

 flora that it is scarcely possible to find precise characters for separating 

 them. Siicli are the leaves of M. pseudo-acuminnta mentioned above, com- 

 pared to those of M. acuminata Linn., the well-known and common cucumber 

 tree of our eastern flora, which is represented in the Southern States by 

 M. cor data Michx., considered by some botanists as a mere varietv of the 

 preceding. The type is recognized, as already said, in leaves collected bv 

 the Princeton expedition from Montana, in those of M. ovalis Lesq. of the 

 Eocene Flora of the Mississippi, and in those of M. cnlifornka of the Upper 

 Miocene of the auriferous gravel deposits of California. It is the same with 

 M. temufoUa, whose leaves are represented in the Dakota Group, being allied 

 by their form and peculiar nervation to those of the living Ma(/i/olia iniihrclla 

 Linn., of the Southern States. 



The genus Liriodendron is represented in the Dakota Group by a large 

 number of leaves, whose characters are so peculiar and so diversely modi- 

 fied that they have been referred to ten different species. The diversity and 

 multiplicity of the leaves have been already remarked ujjon with more 

 details and put in full exidence. After all this, is it not remarkable tliat no 

 remains of plants referable to Liriodendi-on have as yet been oljserved in 

 the Cretaceous of Europe, and none in the Upper Cretaceous of Greenland 

 and of Noith America? And in the Tertiary or more recent geological 

 formations, the genus is recognized only by leaves with \-ariations so little 

 marked that they are all generally considered as referable to a single species. 



