NoVEsrBKE 9, 1900.] 



SCIENCE. 



709 



types in succession ; often a dune phase is 

 interposed in this series, immediately after 

 the beach. 



2. Hydrophytic to Mesophytic. Hydro- 

 phytic areas are common in young regions 

 and are either drained or undrained. Un- 

 drained lakes and swamps are very com- 

 mon at first, but are veiy rapidly filled by 

 vegetation, so that one formation rapidly 

 follows another from the lake to the forest ; 

 zonal arrangement is usually found in these 

 places. Drained swamps and rivers often 

 increase as a region grows older ; progres- 

 sive development is best seen on the flood 

 plains, where the order of succession is 

 commonly well marked and rapid, culmi- 

 nating in the very highest type of meso- 

 phytic forest. There are often hydrophytic 

 shoi-es along the lakes, usually in the less 

 exposed places ; their history is much like 

 that of a swamp. 



B. Retrogressive Development of Plant 

 Formations. Retrogression is commonly 

 local or evanescent. It is best seen along 

 lake or river bluffs, where constant erosion 

 causes the destruction of mesophytic forma- 

 tions. When erosion ceases, progressive 

 movements begin, culminating again in 

 mesophytic floras. Retrogressive move- 

 ments may also be caused by crustal move- 

 ments, changes in climate, or through the 

 action of man. ^^^^^ ^_ ^^^^^^^ 



Univeesity of Chicago. 

 II. 



THE RELATIONS OF THE NORTH AMERICAN 

 FLORA TO THAT OF SOUTH AMERICA. 



In my paper on ' The Relation of the 

 Flora of the Lower Sonoran Zone in North 

 America to the Arid Zones of Chile and 

 Argentine,' attention was called especially 

 to discussions by Gray and Hooker and by 

 Engler on the presence of North American 

 or boreal floral elements in South America. 

 The species considered in the two citations 

 were chiefly alpine and mountain xerophil- 



ous plants of the Rocky Mountain Region 

 and the arid Southwest (the latter especially 

 by Engler) which occur in the Mexican 

 Cordilleras and in the boreal altitudes of 

 the tropical Andes, becoming more generally 

 distributed in the extra-tropical Andes and 

 the higher plains of Chile and Argentine. 

 My own paper attempted to show that a 

 very significant number of the genera rep- 

 resenting the most extremely xerophilous 

 elements of the enclosed desert plateaus 

 and valleys of the Lower Sonoran Zone, 

 reappear in correspondingly arid regions 

 far south of the equator, and that the in- 

 tervening territory contains these rarely or 

 not at all. It further discussed the prob- 

 lems of distribution between the two re- 

 gions, going in some detail into a discus- 

 sion of certain species which illustrate the 

 case especially well. 



In this paper the purpose will be to 

 point out the generally known and accepted 

 facts of relationship between the floras of 

 North and South America as illustrated 

 in all the floral elements represented in 

 both, emphasizing more particularly the 

 elements which I have studied in some de- 

 tail which furnish additional evidence for 

 conclusions already suggested rather than 

 offer a new solution to the more diflQcult 

 problems of distribution. 



It may be said as an elementai-y observa- 

 tion, that if we consider only the present 

 aspects of plant life, and conceive the floral 

 zones of North and South America to be 

 due to and lie coincident with zones of lati- 

 tude, we should have in the two Americas 

 only the tropical zone in common, shading 

 off into the north and south zones of lower 

 temperature, in which the likelihood of a 

 mixture of boreal and austral elements of 

 any two corresponding boreal or austral 

 zones would grow less with increasing 

 proximity to the poles. The question of 

 distribution would be chieflj' one of distance 

 which might or might not be overcome by 



