148 T O R R E Y A 



cases in which it is definitely shown that mutations participate in the origina- 

 tion of geographical races. 



When introgressive hybridization (Anderson and Hubricht, 1938) is dem- 

 onstrable and when a series of chromosome changes, such as a polyploid series, 

 can be shown along highways radiating from a center, it would seem that the 

 indication of center of origin is incontrovertible. When several characters show 

 a parallel and direct continuity of gradation of frequency or of modification, it 

 is likely that there has been active migration of the population from a center. 

 This is sometimes recognizable by chains of subspecies, pairs of species, etc. 

 Payson's (1922) work on LesquereUa provides a good example based on com- 

 parative morphology. He says, "In a graphic representation of the subsectional 

 groups they may be shown by lines radiating from a common center. Such a 

 diagram could be superimposed upon a map and in nearly every case the spe- 

 cies at the base of each line of development would be nearer the Texas region 

 (center of origin) than species derived from it." 



Once again it can be said that this criterion alone is of no significance. A 

 geographic series of size expressions may be due to environmental conditions 

 reflected in growth responses (phenotypic changes in a genotype) or it may be 

 due to selection operating through a region of gradually changing environment. 

 When morphological, phylogenetic, and geographical data are used to support 

 one another, the validity of the conclusions regarding direction of migration 

 depends upon the validity of the morphological criteria employed. 



Criterion 9. Direction Indicated by Geographical Affinities 



This criterion is frequently valid for organisms located at stations removed 

 from the major area they occupy. As mentioned earlier, in any region there are 

 usually numerous extraneous species representing two or more different floris- 

 tic elements, and recording as many different migrations in the vegetational 

 history of the region. In this connection Grinnell and Swarth (1913) say, 

 "We cannot expect to derive universal laws for the behavior of species, to be 

 applicable uniformly in any region . . . where two faunas meet . . . Upon reflec- 

 tion it is difficult to conceive of precisely the same set of delimiting factors oper- 

 ating upon any two species alike." For extraneous species, it is frequently a 

 fairly safe assumption that they were derived from the areas where they have 

 their principal distribution. If a genus or family is largely characteristic of a 

 single formation or climatic type, and has one or a few species of different 

 type, it is likely that the latter migrated and evolved from the generic center. 

 Bromeliads have migrated away from the humid tropics and entered the deserts 

 of southern Mexico, and, conversely, cacti have migrated out of the desert re- 

 gion and established themselves as epiphytes in the tropical forest, according to 

 Gleason (1923). No one suspects certain rather large tropical groups as hav- 



i 



