Morphological Variation and Its Causes in A. tigrlnuni 51 



.angle of less than 45° to the substratum. Even minor degrees 

 of difference in body form and other characters which showed 

 the development in greater or less measure of one or the other 

 habit proved to be correlated, with surprising delicacy, with the 

 special angle at which the sole of the posterior foot naturally 

 rested on the bottom. It should not be supposed that such a rule 

 would hold for larvae taken at random from various sources, 

 and it does not so hold. But in the case cited the differentiating 

 factor between these two groups had been solely their relation to 

 the bottom. As to more conspicuous contrasts, involving, be- 

 sides the angle of the foot, the flexure of all the joints of the 

 limb, they are not usually to be found in young larvae of the first 

 summer. But among neotenic specimens occasionally the strong- 

 est differences appear. Some individuals will show the strong 

 "one-jointed" leg described above, others will vary from this to 

 more and more slender limbs, held in more or less useless atti- 

 tudes, sometimes actually trailing behind, not only when the ani- 

 mal swims freely but when it is undulating slowly along the bot- 

 tom ; and, finally, many of the neotenic larvae will show limbs 

 fully adapted to walking by the strong flexure of all of the joints 

 and the natural plantation of the soles. Such extremes as this 

 last described occur only in larvae that have lived a long time on 

 the bottom. I have observed the development of these different 

 limb-habits to a considerable extent in larvae grown under my 

 observation, but not to the fullest extent in the case of the walk- 

 ing limb. In the other direction, however, I have even seen the 

 posterior limbs of a beautiful larva reduced in size and become 

 utterly useless from no other cause than that the specimen, quite 

 normal at first, was rapidly grown in a tall glass jar where a still 

 stronger companion allowed it absolutely no space at all for limb 

 exercise on the bottom. These extreme differences in the larval 

 limbs are also not obliterated or even reduced by metamorphosis. 

 I have now spoken of variation and causes of variation as I 

 have found them present and operative in the case of several ex- 

 ternal characters of the species. These variations, or some of 

 them, are sufficiently striking in amount, although my chief in- 



247 



