Distribution of the Flora in the Alpine Zone. 49 
In the first area containing 661 species, the agreement of the 
coefficients is surprising: in the second and third it is still very 
remarkable ; and the fact that it manifests itself in such small 
numbers of species (240 and even 92) leads us to consider it as a 
fundamental law of distribution. 
An examination of the floras of extensive areas ( e.g ., Switzerland, 
France, Germany) shows that this agreement is quite general:— 
Generic Coefficients of 
Germany Switzerland France 
2500 sp. 2450 sp. 4250 sp. 
Dialypetalae ... 28 # 3% ... 26-2% ... 19% 
Gamopetalae ... 28 ... 26 ... 19 
Total Flora ... 28 ... 26-4 ... 19-7 
The agreement which occurs, in areas whose extent and 
floristic richness are very different, between the generic coefficients 
of Dialypetalae and Gamopetalae and of the whole flora, seems to 
indicate that the plants of these two classes have a sensibly equivalent 
power of adaptation, since in the competition which takes place among 
them for the conquest of ground, their different genera present a similar 
specific diversity. 
This being so, is it not legitimate to hold that in the struggle 
which takes place between them, individuals are not only the 
representatives of such and such a species more or less well adapted, 
but also the champions of higher biological groups, of genera, 
orders and classes, to which they secure a definite proportion in 
the entire distribution ? 
VI. 
To sum up, analysis permits us to consider tbe distribution of 
plants in the alpine zone as a resultant of the combined action of 
three orders of factors : (1) ecological factors (nature of soil and 
climate) ; (2) biological factors, expressed by the degree of adaptation 
of species to their station, and better still by the power of adaptation, 
which is very unequal in different species ; (3) sociological factors 
created by the competition which occurs between associated species. 
The action of the first two factors has, as a consequence, in 
any station, the elimination of a certain number of species (elimi¬ 
native selection). The third factor determines the local distribution 
of the species which are not eliminated (distributive selection). 
Tbislastkind of selection being at once numerical and taxonomic, 
we may distinguish : (1) A numerical selection, determining the 
number of individuals, and of associated species; (2) A specific 
selection, determining the nature of associated species, i.e. their 
