i 3 o PROBLEMS OF GENETICS 



question not yet answerable with certainty. Coutagne moreover 

 lays stress on the fact that although each species may be easily 

 known from the other in its own district , yet when shells from 

 different districts are brought together it is sometimes impossible 

 to sort them. He mentions an example of such casual inter- 

 mixture occurring under natural conditions on an island in the 

 Rhone, to which it may well be supposed that floods had brought 

 immigrants from miscellaneous localities. This population con- 

 tained a very large number of uncertain specimens, and as he 

 says, it was much as if he were to mix the shells from his 62 local- 

 ities, after which it would certainly be impossible to separate 

 the two species again. 15 



Further evidence is given in the same treatise as to other 

 examples of polymorphism, especially in the genus Anodonta, of 

 which Locard made 251 species for France alone. Here again 

 are cases like those already given, and many forms or "modes" 

 are found restricted to special localities, while occasionally 

 in the same locality dissimilar forms are found, collectively 

 forming a colony, without intermediates. 



Taken as a whole the evidence shows the following conclusions 

 to be true. Local races, whether of animals or plants, may be 

 distinguished by characters which we are compelled to regard 

 as trivial, or again by features of such magnitude that if they 

 were known to us only as the characteristics of a uniform species 

 they would certainly be assumed without hesitation to be essential 

 for its maintenance. Local forms may be sharply differentiated 

 from the corresponding populations of other localities or they 

 may be connected with them by numbers of intermediates. 

 Not rarely also we find a fact which has always seemed to me of 

 special significance, that the peculiarity of the local population 

 or colony may show itself in a special liability to variation, and 

 this variability may show itself in one of many degrees, either 

 in the constant possession of a definite aberration, in a dimor- 

 phism, or in an extreme polymorphism. 



At this stage attention should be called to two points. 



15 With this evidence compare that given by A. Delcourt in his valuable 

 papers lately published relating to the variations of Notonecta. See especially 

 Bull. Sci. Fr. Belg., 1909, XLIII, p. 443; and C. R. Soc. Biol, 1909, LXVI, p. 589. 



