Mor[>hological I'ariatioii and Its Causes in A. tigriniiin 73 



* 



Lastly, the foregoing presentation raises the general question : 

 What is the significance for evolution of such variations as A. 

 tigrinnui shows us? Are such variations the stuff of which spe- 

 cies are made? This may well be doubted. Not all species, as- 

 suredly, arise thus. Did they do so, they would be as temporary 

 as the temporary local conditions of environment that produced 

 them. They would vary much more than they do ; and instead 

 of the broad facts of geographical variation we should find, in 

 the most limited ranges, a chaos now of minor, now of major 

 differences, all determined by the most purely local conditions. 

 It is, indeed, evident that, with forms as plastic as A. tigrinum, 

 the formation of a new species, in the ordinary sense of the word, 

 must imply a cessation of variation rather than its continuance. 

 Otherwise the new form will be as plastic as the old and will be 

 reduced by the environment to a common level, or a common 

 chaos, with it. The only way that such variation as that shown 

 by the species here discussed could be productive in species for- 

 mation would be through the control of environment itself. 

 Thoroughgoing, broad, and general differences in environment 

 would readily stamp the individuals produced under them with a 

 sufficiently peculiar character. The result would be an "onto- 

 genetic" species. That such species exist, both in nature and in 

 our catalogues, seems almost certain. But to discover them and 

 test their character fully will require experimental studies along 

 lines which I hope this article may, to some extent, suggest. If 

 botanists are discovering that the only test of a species is in the 

 transplanting, it seems not improbable that the zoologist must 

 soon admit that the final test of many species must lie at least in 

 the rearing, and that, too, under controlled conditions. 



But even the discovery and proper classification of ontogenetic 

 forms is of less importance than the tracing of their relationship 

 to forms that are incipiently phylogenetic. The genus Ambly- 

 stoma is rich in material for such study, both within as well as 

 without the species A. tigrimun. 



In closing, I wish to thank Dr. H. L. Shantz and Messrs. R. M. 

 Ragle and M. G. Hall for the service they have rendered me in 

 securing very valuable material from Colorado. And finally, I 



269 



