Morphological J'ariation and Its Causes in A. tigriniun 23 



involve keeping them for a good portion of the time in water. 

 Last season, however, I was able so to situate thirty adults in 

 their third summer that they ate prodigiously and yet were out 

 of the w^ater, in a fairly moist atmosphere, for the greater pro- 

 portion of the time, and when in the water thev practically never 

 swam, but only crawled lazily about. The tails of many became 

 twisted permanently to one side under these circumstances. Buj; 

 this did not prevent the customary hypertrophy of a number of 

 specimens which showed broad and thick, or corrugated thin 

 tails according to sex. The broad tail thus need not be a highly 

 functional swimming organ. Eighth, and finally, the chief point 

 which I wish to emphasize in regard to this organ is that, how- 

 ever much it may be modified in the adult (and by artificial means 

 this modification may become prodigious), its customary varia- 

 tions are not due to this means, but, like all the chief variations in 

 this species, are preformed in the larva. Not only the tail fin of 

 the larva, but the tail as well, varies very greatly in width. A 

 glance at plates I and II will show something of this, although no 

 slender-tailed types are there shown from the side. A comparison 

 of figure I, plate I, with figure 2, plate III, will also show to what 

 extent the tail of the adult resembles that of the larva minus the 

 finfold. A slight reduction in the robustness of the tail does take 

 place, differing in different individuals according to the state of 

 nutrition and the circumstances of the metamorphosis, but these 

 changes are unimportant and relatively uniform compared with 

 the very considerable variations in form of the organ due to the 

 accumulated influences of larval life. I have been unable to figure 

 any extremely long and slender tailed adults for want of space. 

 But a comparison of figures i and 2 in plate III is interesting, 

 especially when the history of the two animals is known. Both 

 specimens owe their peculiar tail-form to larval development, and 

 not to changes subsequent to metamorphosis. The short, broad- 

 tailed adult (which is, I may say, a more extreme animal than the 

 other) was photographed a few^ weeks after the completion of 

 metamorphosis. Moreover, it proved to be very constant in form, 

 showing surprisingly little change after seven months' terrestrial 

 life in a rather warm basement, or again after a summer's heavy 



219 



