164 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



their allies among the Compositae; and a great number of others. 

 There are over seventy of these groups among the Dicotyledons 

 alone. The only woody genera which display such a segregation 

 are Cercocarpiis and its allies, A\hich comprise the subfamily Cer- 

 cocarpeae of the Rosaceae, and number only a very few species. 



A similar constitution of the endemic flora may be observed in 

 Europe where, centering in the Mediterranean region, there are 

 scores of endemic herbaceous genera almost all of which fall into 

 sharply marked groups of closely related forms. This is par- 

 ticularly noticeable among the Cruciferae, Leguminosae, Um- 

 belliferae, and Compositae. 



These two great arrays of what we have called "indigenous" 

 endemic genera in North America and Europe seem to have their 

 centers of dispersal in the western and southwestern United States 

 and northern Mexico on the one hand, and in the Mediterranean 

 region on the other. That they are confined almost entirely each 

 to its own side of the ocean; that both are composed of groups of 

 closely related genera, very many of which are rich in species; 

 and that each constitutes a dominant and successful element in 

 the flora of its own region, all suggest that such endemic types 

 have either actually had their origin, or have at least undergone 

 the greater part of their development and dispersal since a free 

 exchange of plants between Europe and North America was dis- 

 continued, presumably somewhere about the middle of the 

 Tertiary. Had they been as common and widespread before that 

 time as today, they would in all probability be represented now 

 on both sides of the Atlantic, as are so many genera. 



We may feel justified in concluding, therefore, that the "relict" 

 and the "indigenous" endemics in the north temperate zone do 

 indeed diff"cr considerably in the degree of their antiquity; the 

 former representing decidedly ancient types, which may be held 

 to correspond to the ancient endemic genera of other regions, but 

 the latter seeming to have arisen in much more recent times. 

 The significant fact in the whole matter is that, with very few 

 exceptions, the indigenous endemics are composed entirely of 

 herbaceous forms. \n this wholesale development of new generic 

 types, why should not trees and shrubs have taken the place which 

 their abundance as species and individuals would seem to warrant? 

 Why have not our local varieties and species of woody forms, 



