Morphological Variation and Its Causes in A. tigrinuni 37 



as I find by repeated experiment, always greatly reduces the size 

 of the gills (see fig. 3, pi. VI). 



The question remains, although the answer has already been 

 partially indicated, as to how far these special forms of larval 

 head are transmitted to the adult. The general reply is that they 

 are all retained to a greater or less extent, in so far at least that 

 the young adults, after metamorphosis, show in a partial if re- 

 duced degree the form of head acquired by the larva. Even the 

 delicate character of the truncated muzzle is by no means oblit- 

 erated. Some features are obliterated m.ore, however, than others, 

 e. g. extra width of head is lost more than is extra length. More- 

 over, the nature of the circumstances of the metamorphosis also 

 conditions the result. Slow metamorphosis at low temperatures 

 absorbs more of the larval tissues, especially if, as is usually the 

 case under such circumstances, no food is taken during the 

 process. Rapid metamorphosis at higher temperatures usually 

 absorbs less, and especially if, as I find to be frequently the case, 

 the animal does not entirely discontinue feeding for more than 

 one or two days of the metamorphic period. Thus slow and 

 forced metamorphosis reduces the width of the larval head more 

 than does rapid metamorphosis or metamorphosis plus some feed- 

 ing. Or again, and this proves a striking point, slow and sheer 

 starvation-metamorphosis produces an adult with a much wider 

 gape of the mouth than does metamorphosis of an opposite or at 

 least less extreme nature. 



Both of these results happen to be exactly the opposite from 

 what we should have expected from Cope's assumptions. As an- 

 other example. I may mention that metamorphosis out of water 

 has most peculiar results, although these pertain less to the head 

 than to other parts of the animal. And, finally, there is one inter- 

 esting character, never to my knowledge possessed by the larva, 

 and very rare among adults, that is produced entirely, in my ex- 

 perience, by difficult or retarded metamorphosis. This is the pro- 

 longed or protruding lower jaw. It is a chief character in the 

 diagnosis of one of Cope's one-specimen species, viz., A. xiphias. 

 I do not recall ever having seen this feature in an adult resulting 

 from metamorphosis during the period from early to late sum- 



233 



