2.6 /. H. Powers 



thus appears as if partially disjointed from the body, instead of 

 constituting an evenly tapering prolongation of it. Even this 

 gross thickening of the larval tail, which remains so conspicuous 

 in the young adult, proves to have a perfectly definite and recog- 

 nizable cause. It is again one of nutrition and activity. When 

 larvae which have reached a considerable size have their nutrition 

 slowly checked there results an equally gradual shrinking of the 

 back and tail fins. The area of this swimming organ may be eas- 

 ily reduced to less than half its former size before metamorphosis 

 really intervenes. This always happens with wintered larvae. 

 When nutrition is again increased the fin surface may reexpand 

 or it may not. Seldom does it reexpand to its former proportions, 

 although the animal may grow meanwhile to twice its former 

 bulk. With every vicissitude in nutrition, therefore, the swim- 

 ming surface of the tail tends to be more or less permanently re- 

 duced relatively to the weight of the animal. If, now, the larva 

 with reduced tail surface is obliged to swim from habit or neces- 

 sity, there results a very marked increase in the development of 

 the posterior portion of the body, including the tail. The narrow- 

 finned larva makes up in strength what it lacks in surface. I have 

 often verified these results, in aquaria as well as in nature. 



VARIATIONS SHOWN BY THE HEAD 



Here we find at first a bewildering number of types and indi- 

 vidual differences. Of the extent of variation in the breadth of 

 the head relative to the length of the animal I have already spo- 

 ken. But its other dimensions are no less variable ; its length, its 

 thickness, its contour, the rounded, tapering, or truncated muzzle, 

 the development of the so called parotid glands, the distance be- 

 tween the orbits and the nares, and especially the gape of the 

 mouth, — all these are quite as variable as is the width of the head. 



To disentangle the factors which might produce these forms 

 seemed at first hopeless. But in the course of over a hundred 

 separate experiments in raising larvae to and beyond the period 

 of metamorphosis, although these experiments were made with 

 other ends in view, vet as conditions were controlled and varied 



