28 J. H. Powers 



that they show an extreme plasticity and variation in develop- 

 ment. The parotids in figure i, plate II, are really much thicker 

 than they seem, making up a considerable proportion of the width 

 of the head. In figure 2, plate II, it is rather the sideward ex- 

 tension of the large branchial apparatus that broadens the head, 

 although the parotids are here, too, heavy. Excessively protrud- 

 ing parotids like those in figure i are essentially phenomena of 

 overnutrition. Out of hundreds of larvae taken during the same 

 season from the same pond not a specimen showed the protruding 

 cheeks during the summer. But after the great multiplication of 

 daphnids' in the early fall, a few weeks sufficed to transform many 

 of the larvae in the pond into the likeness of figure 1. They were 

 feeding very heavily in nearly clear water and were resting mainly 

 upon the bottom. I have observed the same thing over and over 

 again. 



Low nutrition, on the contrary, or starvation at certain periods 

 of larval life so reduces these parotid regions that, unless the gape 

 of mouth is very narrow, the sides of the head become nearly or 

 sometimes quite parallel. I have produced a great many such 

 larvae experimentally, and have found nearly similar specimens 

 in nature under appropriate conditions. Specimens from the 

 Seven Lakes, Colorado, where the species is said not to meta- 

 morphose, and where sexual maturity is attained at a very small 

 size, show this same parallel-sided head, together wuth many 

 other signs of the lowest nutrition. 



Another result of low nutrition, and especially of periods of 

 sheer starvation, consists in the great flattening of the head. The 

 starvation of large, normal larvae in the fall, the only period at 

 which this can be easily done without producing metamorphosis, 

 often has very striking results of this nature. Such results are 

 erratic, not affecting equally all similar specimens exposed to the 

 same treatment ; but some respond to it quickly, and, like so many 

 other effects of transient nutritive causes, this result of starvation 

 is frequently quite permanent. It does not succumb to later pe- 

 riods of full feeding, and is fully transmitted to the adult. This 

 variation is not well shown on the plates, although figures 4 and 5 

 on plate II, taken from the same animal, show something of it; 



224 



