T2 J. H. Poivcrs 



1 



of typical build, being the result of the free-svviniming habit with 

 abundant food. Figure 2 of the same plate represents a corre- 

 sponding larva, a little more slender, from another pond where 

 food was less abundant. It will be noticed that the tail fins of the 

 two compact larvae are much mutilated, while the more slender 

 specimen from the same pond has but a single mutilation. This 

 contrasting condition was invariable with the two types, save that 

 many of the free swimmers showed no mutilation. This condition 

 was due to the numerous crayfish which lay in wait for the larvae 

 at the bottom but which could not reach the others. Other proofs 

 of their differing life habits might be given. In regard to these 

 special larvae I shall recur to one such proof in speaking of varia- 

 tion in the case of the posterior limbs. I may say, however, that 

 as to the existence and persistence of the types I have had abun- 

 dant opportunity for observation mider conditions that permitted 

 of no doubt. Even in aquaria the differences in habit as well as 

 structure become very marked. The lazy, heavy-bodied animal, 

 wanting oxygen, will raise the head and start slightly upward, 

 relapse again into quiescence, again start forward, and yet even a 

 third time, before it finally raises itself unwillingly to the surface, 

 to take in, when it finally gets there, an enormous gulp of air. On 

 the other hand, the swimmer never waits for a second impulse 

 when its respiratory period (the animals visually rise at regular 

 intervals) is completed, and not infrequently comes close to the 

 surface as if for oxygen when no real need has yet arrived. 



I have made considerable effort to ascertain what the causes 

 might be which lead the animals thus to differentiate their habits 

 of locomotion, feeding, etc. Sufficiently precise observations are 

 not possible in the field. And in aquaria often no cause whatever 

 can be assigned. The animals somehow seem different. I have a 

 little evidence that hereditary predisposition is sometimes at the 

 basis of the incipient differences, which rapidly gain in force as 

 habit becomes fixed. 



In other instances, however, a perfectly evident cause does ex- 

 ist for the assumption of divergent locomotor habits, these latter 

 being, it would seem, forced upon the animal by a sudden and 

 rapid change in the form of its development. A weak-limbed, 



208 



