Cos 



THE AMERICAN NATURALIST [Vol. XL1II 



then, may be sharply demarcated despite of the possibility 

 of reciprocal derivation. 



Lastly I had wished to defend the seeming paradox that 

 species may be not only real, but all but absolutely stable 

 despite of the widest variability. This is not a contradic- 

 tion in terms. The stability of a species depends upon 

 its refusal to vary in certain given directions, i. e., away 

 from its specific characters, or, secondly, upon the non- 

 transmissibility of such variations when once produced. 

 The variability of a species, however pronounced, may 

 mean only the production of non-specific characters or 

 the production of characters of whatsoever order which 

 are not repeated in the offspring. 



Thus it proved to be with the species of salamander— 

 Ambly stoma tigrinum— wpon which I spent several years 

 of almost continuous experimentation. Variability and 

 instability of species, when I began my work, for me 

 were synonyms; when I concluded they had lost almost 

 all relation in meaning. The astounding variation of this 

 species was in the main but a somatic by-play in re- 

 sponse to environing forces. However, wholly against 

 first impressions, it turned out that this somatic variation, 

 despite its variety and extent, yet had its marked limi- 

 tations. How I did strive to make Amblystoma punc- 

 tatum out of A. tigrinum— such a little thing, too-just to 

 make a leopard salamander out of a tiger salamander. I 

 did not even try it until my third season's experimenting. 

 It was really too insignificant a task. Had I not observed 

 nature working much greater changes ?-and, by imi- 

 tating her methods in experiment, had I not gained the 

 key to her processes? Had I not passed the bounds of 

 specific and even generic characters ? Indeed, certain wise 

 ones had nodded gravely, and suggested no less than f am- 

 ■ly rank for the best of my handiwork. Just to make A. 

 Vunctatum out of A. tigrinum! Besides I had the thing 

 three quarters done already, time and again, as a by- 

 product of my other work. At last I concluded, however, 

 to make a few bona fide A. punctatum just for the fun, 

 and to plague certain species people. And for season 



