( 149 ) 
development or merely lateral branches of the Eavernonieae line. 
In the Lychnophoreae one line begins with Hremanthus and is 
characterized by the spirally arranged scales of the involucre. The 
species of this group are entirely South American. A second 
includes the Elephantopeae, in which the scales are decussate, and 
the third leads to the genera Rolandra and Spiracantha, in which 
the corolla is merely toothed, and the style-branches are short. 
Within these series the genera are distinguished by the structure 
of the pappus, which may be scaly, of scales and bristles, entirely 
of bristles in one or more similar series or two dissimilar series, or 
certain bristles may be peculiarly modified; by the number o 
flowers in each head, which is found to be constant for each spe- 
cies within certain limits; and by the character of the inflorescence. 
In 1898 there appeared a paper by Wettstein* in which the 
author presented in a very convincing manner the insufficiency of 
orphological comparison alone in studying the relationships and 
probable phylogeny of species. He introduced examples showing 
how conceptions of species may be changed by combining distribu- 
tional with morphological studies, and his own results, particularly 
on the genera Gentiana and Huphrasta, are proof of the value of 
his method. Stated in a somewhat modified form, his principle is 
that in a polymorphic genus the most nearly related species will 
occupy the same or adjacent areas, or will at least be connected by 
definite migration routes. In studying the relationships of the 
species of Vernonieae in North America this geographical method 
has been used as far as possible, and while in but few cases it led to 
conclusions which would not have been reached by morphological 
comparison alone, it has invariably confirmed the results of the 
latter in a most striking way. 
Naturally an abundant supply of herbarium material or of field 
notes is necessary for successful application of Wettstein’s method, 
so that the range of species, both geographically and ecologically, 
and their variation can be definitely ascertained. It is also impor- 
tant that the general phytogeographical relationships should be 
known, with some accuracy, of the floras in which the species to be 
investigated are found. Under phytogeographical relationships are 
included such features as centers of distribution, migration routes, 
Resi emaass der geographisch-morphologischen Methode der Pflanzen 
systematik 
