127 



known, three belong to the middle zone of the United States, which has also 

 for its share all the living species of Carya. Of the living species of Rhus, 

 Austral Europe has two, one of which, R. cotinus, has been compared, by the 

 form of its leaves, to Bumelia emarginata of the Dakota group. We have in 

 the United States, beside Rhus metopium, which is rather a tropical form, six 

 species of the section of the pinnately-divided leaves, with the trifoliate R. 

 toxicodendron and R. aro?natica, both extremely variable, all types already 

 represented in the Pliocene of California. 



Except an Amelanchier, described by Dr. Newberry in his Notes on 

 Extinct Floras, &c, from the Tertiary beds of the Yellowstone, we do not 

 know as yet any fossil species of Rosacea, from the western Tertiary measures 

 This is not a reason why Prunus should be excluded from the list of the 

 genera of the Dakota group. By their present distribution, our P. serotina 

 and P. virginiana indicate an extreme power of life, or of resistance to climatic 

 changes, both being the only arborescent species of this continent having a 

 range of distribution of thirty to thirty-five degrees in latitude, and both, too, 

 being found everywhere, on every kind of ground ; the one as a shrub along the 

 banks of streams, the other as a fine tree in our woods. And, also, we have in 

 our P. caroliniana, a shore-tree of the South, a species whose coriaceous, entire 

 leaves recall the essential characters of those of the Dakota group. Three 

 species of this genus are described from the Tertiary of Europe, and none as yet 

 from ours ; but it is probable that fossil remains referable to it will be found 

 hereafter, as it has in our present flora a larger number of species than in that of 

 Europe, or of any other part of the world. Of the species described by De 

 Candolle, fourteen are North American, five European, four species belong 

 to Japan, &c. 



Resuming, in a few sentences, the above remarks, we find that the dico- 

 tyledonous flora of the Dakota group represents species referable to the genera 

 Liquidambar, Populus, Salix, Betula, Myrica, Celtis, Quercus, (in two of its 

 principal types,) Ficus, Platanus, Laurus, Sassafras, Cinnamomum, Diospiros, 

 Aralia, Magnolia, Liriodendron, Menispermum, Negundo or Acer (?), Paliurus, 

 Rhus or Juglans^), and Prunus (I); or, merely considering the affinities to 

 our present flora, of twenty-one genera, seventeen of which are those to which 

 belong the species of our trees and shrubs which have the more general and 

 the widest range of distribution. Indeed, all our essential arborescent types 

 are there, except those which are marked by serrate or doubly serrate leaves: 



