117 



land-plants of the Eocene, especially none of its essential representatives, 

 the Palms. The proportion of Palms, especially of Sabal species is marked 

 in the lignitic at Golden, Black Butte, &c, not only by the remains of leaves, 

 which in places fill thick strata of sandy clay, but also by fossil wood of the 

 same class of plants, or by their trunks transformed into coal and identified by 

 the characters of their internal structure. In ascending from the lower 

 lignitic measures, where the essential types of the Cretaceous flora have no 

 representatives, we see these Cretaceous types re-appearing, a few in the 

 Upper Eocene of Evanston, more of them in the Carbon group above, still 

 more in the Upper Tertiary, following thus an increasing degree of predomi- 

 nance, culminating, it seems, at the present time, in the flora of the eastern 

 slope of the North American continent. The disconnection of types of the 

 flora of the Dakota group appears, therefore, as a kind of break in the vege- 

 table scale, accountable perhaps to modifications of climatic circumstances. 



The essential and more numerous vegetable remains in the Dakota group 

 are leaves of Dicotyledonous, representing the three divisions of this class, and, 

 what is more remarkable, the genera to which belong most of the living 

 arborescent plants of this country and of our present climate. If what may 

 be called positive characters of the genera — the flowers and the fruits — are 

 not ascertainable from fossil remains, it is at least imjjossible to deny the 

 intimate relation of most of the leaves of the Dakota group to the genera to 

 which they have been referred in their descriptions. 



Begin»ing by the Apetalous, we have, first, Liquidambnr leaves, so simi- 

 lar to those of our sweet-gum tree (L. styracrfluuni) by form and nervation, 

 that, in comparing the fossil leaves with those of our living species, no other 

 difference can be remarked but the entire borders of the fossil ones. 

 They are more or less serrate-crenulate in the living species, as also in L. 

 europeiim of the Miocene of Europe. But some sjjecies of the same forma- 

 tion, considered by authors as referable to this genus, have leaves with entire 

 borders, as seen in the descriptions. Even Gaudin, in his first Memoire 

 on the Fossil Leaves of Tuscany, figures as L. europcum, three leaves, one 

 of which, with entire borders, (PI. v, Fig. 3,) is remarkably similar to our 

 Fig. 1 of PI. ii ; the lateral nerves being marked as branches of the second 

 pair of nerves, just as it is in our Cretaceous leaves, and not emerging from 

 the top of the petiole as in the leaves figured by Heer under the same name. 

 Gaudin accounts for the entire borders of this leaf by the supposition that the 



