72 J. H. Poivcrs 



both its general form and the details of its structure, directly and 

 indirectly, as by some new type of locomotion which is forced 

 upon the animal. 



A third general fact of hardly less prominence is the frequency, 

 the variety, and the extent of functional variations. Use and dis- 

 use, in varying modes, degrees, and successions, play strange 

 pranks with this species. Although the direct effects of nutrition 

 are, on the whole, more obvious and perhaps more general as 

 sources of modification, yet such extreme results as "cannibalis- 

 tic teeth" might lead one to consider functional variation as the 

 chief cause of the polymorphism of the species. I am unac- 

 quainted with any instance in which such variation is carried 

 farther than in the larvae of this species. Here in this general 

 principle it is evident that one of the two chief assumptions of 

 Baird and Cope is, in principle, fully justified, although it is a 

 curious conmient on the habit of easy a priori interpretation to 

 find that scarcely one of their applications of the principle holds 

 good. 



A fourth point worthy of special emphasis is this : specific char- 

 acters, in species which vary as A. tigriiimvi varies, are, after all, 

 strongly determined by environing conditions. There is nothing 

 new in this. But the study of this species seems to me to lend it 

 new weight and confirmation. If the broad head and large teeth 

 of the cannibal are acquired characters — and they conform to 

 the definition of such — what are the narrow head and smaller 

 teeth of the customary daphnid-feeder? Are these specific and 

 congenital characters ? They are more frequent, more "typical" 

 in the species ; but I am forced to conclude that they are so chiefly 

 because daphnids are numerous and constitute a convenient and 

 stimulating food. And the same may be said of nearly all specific 

 characters : so readily are they modified by a changed environ- 

 ment that we must conclude they are, in reality, equally deter- 

 mined even by an unchanged environment. Congenital tendencies 

 in such species are not definitely specific, but only indefinitely 

 specific. In this species, indeed, they arc not always even defi- 

 nitelv g'eneric. 



268 



