Morphological l^ariation and Its Causes in A. tigrinum 21 



fully as possible how aquatic the species naturally is in my own 

 vicinity. To sum up in a few words the results I have reached, 

 Baird's assumption is true in so far that it proves possible to de- 

 velop broad, compressed tails in some adults by high feeding in 

 water. The variation thus induced even becomes in a few cases 

 quite remarkable, — see figures i, 2, and 3, plate V. And again, 

 life in a burrow for even a single winter may considerably reduce 

 the breadth of the tail, while old females, kept for several years 

 in moist earth, without access to water, show tails that are almost 

 cylindrical at the base. Rapidly growing summer larvae, too, 

 when forcibly induced to undergo metamorphosis out of the 

 water, develop into adults Avith tails very unlike those of the spe- 

 cies. They are short, cylindrical, and taper rapidly to a sharp 

 point. They resemble the tails of A. opacum rather than of A. 

 tigrinum. 



Striking as the above facts are, however, and confirmatory of 

 a considerable measure of truth in Cope's and Baird's hypothesis, 

 they are not a sufficient explanation of even the major part of the 

 phenomena of tail variation. This is shown by the following con- 

 siderations : first, young adult amblystomas, taken immediately 

 after metamorphosis and kept for three years under the same 

 conditions as to food, temperature, aquatic and terrestrial condi- 

 tions, do not develop like forms of tails. Second, most adult in- 

 dividuals, despite such fluctuations as I have mentioned above, 

 maintain a fairly constant type of tail from year to year, provided 

 they are kept under approximately normal conditions. Third, the 

 broadening of the tails by enforced aquatic life of the adult is, of 

 all the varied modifications that may be produced under experi- 

 ment, the most erratic and uncertain. In a few young specimens, 

 just after metamorphosis, it may take place in a few weeks : others 

 under the same conditions show no effects after months or even 

 years, while now and then an individual that has shown no change 

 during a year of aquatic life will suddenly show an eccentric 

 caudal development. Fourth, under experiment the tails of fe- 

 males seem, on the whole, more easily modifiable than those of 

 males, and the modification may become very marked, — see figure 

 I, plate V. In nature I have never found a female showing more 



217 



