56 



primis, the chestnut oak; Q. ellsivorthiamis and Q. anceps, types of 

 our Q. phellos and Q. imhricaria, species with entire borders of leaves. 



It would be hazardous to pursue further a typical comparison of the 

 Cretaceous species of oaks, on account of the few materials found as 

 representatives of this genus in the shales of the Dakota group. The 

 few specimens, however, represent distinct, well-preserved leaves, from 

 which at least we know that the oaks were already present in the Creta- 

 ceous flora of our continent. They appear few, in a modest way, though 

 already of two distinct types ; but soon the forms become more numer- 

 ous and the genus takes an important place in the arborescent vegeta- 

 tion of the world. In the Eocene flora of the Rocky Mountains six spe- 

 cies have been discovered already, among which one representing the 

 third essential type of our oaks, marked with deeply pinnately-lobed 

 leaves, as in the numerous species of the section of the North American 

 black oaks. The Elk Creek specimens, which seem to represent two 

 horizons of the Tertiary, have eight species; the Washakie group, and 

 Carbon have six, and in the Pliocene of California the representatives 

 of this genus are still more numerous and their types still more inti- 

 mately related to those of the living species. The flora of the California 

 chalk bluffs has six species of oaks, under only thirty-four species. 



The three last genera of the Apetalese represented in the flora of the 

 Dakota group are Platanus, Laurus, and Sassafras. Though no fruit of 

 Platanus has been found till now with the leaves, these are, by their form 

 and nervation, positively typified as representatives of this genus. Heer 

 had already recognized P. JSfewherrii, in his Fhyllites du Nebraslca. To 

 this I have added P. Eeerii, far different from tjie former, as seen in the 

 description ; and P. primceva, which, from its likeness to P. aceroides, I was 

 formerly induced to consider as a mere variety. Though, from the form 

 of other mostly entire leaves, the Cretaceous species is apparently dis- 

 tinct, the analogy or similarity, as indicated by the characters of the 

 leaves, is not the less remarkable. It is the type of the species later rep- 

 resented by acutely lobed and dentate leaves, which we recognize in the 

 Eocene of the Rocky Mountains as P. Saydenii, in the Miocene of the 

 same country and of Europe as P. aceroides^ in the Pliocene of Califor- 

 nia as P. disseetus, and especially now as P. occidentalis. — P. aceroides 

 was already considered by European authors as the ancestor of our P. 

 occidentalis before the Cretaceous species had been discovered. jSTow, we 

 have to refer the origin of our noble tree to a more ancient epoch. 



Like that of Fag us and Liquidambar, the Cretaceous type of Platanus 

 has not widely varied and multiplied, and also it does not appear to have 

 changed its habitat in a marked degree, at least not in latitude. One 

 species only, P. aceroides, and its variety, P. guillelmcB, is abundantly dis- 

 tributed in the Miocene of Europe, from Greenland as far south as North 

 Italy, over an area of about twenty-six degrees of latitude, while the 

 range of P. occidentalis is from the great lakes to the Gulf of Mexico, 

 passing still farther south into Mexico by its analogous P. Mexicana. 

 From Europe it has passed eastward as P. orientalis, in the same way 

 as it has gone west from our country as represented by P. racemosa of 

 California. 



In the Laurinea; we have leaves referable by their form and nervation 

 to the genus Laurus or Persea, and a well-preserved fruit, Laurus macro- 

 carpa, which, comparable also to the fruits of Ginnamomum and Sassafras, 

 is, from its association in the same localities with leaves of Laurus, admit- 

 ted as belonging to this genus. It seems a southern type in comparing 

 it to the other species of the Dakota group, but it is rather, I think, a 

 shore type. Our Laurus (PerseaJ caroliniana extends in following the 



