324 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



just as distinct. They are merely obscured by special conditions 

 which obtain among plants. 



Crossing the Temperate Zone anywhere on east and west lines, 

 we find plant after plant replaced across barriers by closely related 

 forms. Illustrations may be taken from among the higher types— 

 equally well no doubt from among lower ones. Some genera belt 

 the earth or come very near doing so, each form having its twin 

 as next neighbor. A single example, that of the plane tree, 

 Platanus, may suffice. xVnother would be the blackberry, Ruhus 

 villosus, and related species. 



A natural law is not invalidated by the presence of effects due to 

 other causes in the same environment. Actual conditions in nature 

 are everywhere products not of single and simple forces, but re- 

 sultants of many causative influences, often operative through the 

 long course of ages. 



As a rule, related species in almost every group are connected by 

 a fringe of intergradations or subspecies. Where barriers are sharply 

 defined, geminate species also are sharply defined. Where diffuse, 

 geographical subspecies connect them, either wholly or in part. We 

 recognize no difference between species and subspecies except that 

 involved in sharpness of definition. If the particular barrier can 

 not now be crossed, a species resultant from the presence of the barrier 

 will be well defined and therefore unquestionable, however small the 

 elements of difference. Subspecies are almost invariably associated 

 with some feature of geographical distribution. 



It has been claimed that geminate forms are not true species be- 

 cause they often intergrade one with another, and would probably 

 be lost by intermingling were the barriers removed. Some maintain 

 also that only physiological tests can be trusted, as true species will 

 not blend and their hybrids, if they exist, are sterile. This assump- 

 tion is purely hypothetical. Interbreeding is no test of species. 

 Closely related species in almost any group of plants or animals can 

 usually be readily crossed. As the relation becomes less intimate, 

 we find partial sterility of varying grades and at last total incom- 

 patibility. 



In most groups, probably in all, the characters which distinguish 

 species from one another are elements neither useful nor injurious. 

 Unless we take " natural selection " to cover both processes, as 

 Darwin certainly did, we must assign to "selection" the preserva- 

 tion and intensification of adaptive characters, and to " isolation " 

 the seizing and fixing of the nonuseful — usually fluctuating — ele- 

 ment. It is a fact well known to breeders that these indifferent 

 or nonuseful characters are generally more persistent in heredity 



