THB CANADIAN ENTOMOLOGIST 123 



indicate without question the mutation of factors or elements in the 

 germ-plasma. 

 2. Variation due to new combinations, or to recombinations of genes normally 

 carried by a species (as in sex-dimorphism, polymorphism, etc.)- This 

 is the kind of variation w'e are most familiar with, and which has been 

 the subject of so much experimental research. Not only may genes or 

 determiners be shuffled in inheritance, as Mendel described, but those 

 constantly present may be greatly modified by the unexpected appear- 

 ance of others, which have until then escaped notice. Thus in the sun- 

 flower there is a series of pattern-factors, which only become evident 

 when the factor for red rays enters the combination. 



In the case of the many mutants of Drosophila observed by Morgan, 

 all within the limits of a single species, it is difficult to resist the con- 

 clusion that a process of subtraction is going on under our eyes, leaving 

 combinations which are new in the sense of lacking some of the original 

 elements. In other analogous cases, we are struck by the fact that the 

 same kinds of subtractions occur in many different genera and species, 

 showing that the tendency to perform these tricks is deep-seated in the 

 protoplasm of the whole race. There is here involved a question which 

 cannot be said to be settled, and to which much additional research 

 must be directed. 

 3. Variation due to the direct action of the environment, which, as we have 

 learned from Weismann, is not inherited. Nevertheless, the power to 

 react to the environment in particular ways is inherited, and hence even 

 these variations cannot be dissociated from the question of heredity.. 

 On the negative side, as it were, we have the facts of palaeontology. The 

 study of fossil insects shows us that many apparently trivial characters, such as 

 the arrangements of spots on wings, are of enormous antiquity. Not only this, 

 but as Wheeler's researches on the ants of Baltic amber have shown, specialized 

 habits and reactions are likewise millions of years old. It, therefore, becomes 

 more probable that the phenomena of variation which we witness to-day repre- 

 sent, mainly at least, the shuffling of very ancient cards. 



In the Canadian fauna, there are several notably attractive opportunities 

 for the study of variation. I will refer now only to a single family of insects, 

 with which I have been especially concerned in recent years; the Saturniidae. 

 The publication of Packard's Monograph (Memoirs National Academy of 

 Sciences, vol. XII) brings the subject fairly up to date, and makes it convenient 

 to go forward from the point there attained. The genus Samia, as represented 

 in Canada, is extremely interesting. There is not only the question of the 

 relationships of the species cecropia, gloveri, Columbia and rubra, but Columbia 

 has in the west a remarkable race nokomis (Brodie), while rubra produces at 

 Kaslo a form kasloensis Ckll. These are merely conspicuous outstanding facts; 

 large collections from many localities, together with experimental breeding, 

 will bring out innumerable details of interest. Another very interesting species 

 is Telea polyphemus, the variations of which should be studied exhaustively, 

 and compared with those of the great Asiatic silk-moths. The Hemileucidae, 

 also included in Packard's book, afford similar opportunities, and owing to 

 their smaller size are somewhat more manageable. 



