Morphological Variation and Its Causes in A. tigriniim 7 



cause it is, as will be shown, subject to the most independent and 

 extreme variations. All in all, however, the width of the, head 

 is the most available dimension to use as a measure of the animal's 

 robustness. It is easily measured, and, despite notable devia- 

 tions, there is a more or less general correlation between its width 

 and that of the hipgirdle, etc. 



Taking, then, as a most general expression of bodily proportion, 

 the ratio of the width of the head to the total length of the ani- 

 mal, we find a general variability in the species that is very great. 

 Cope records measurements for the species, showing the width 

 ,of the head entering the total length from 7.88 times to 10.78 

 times. This no doubt includes sex differentiation (the mature 

 males being, as I shall show later, proportionately longer tailed 

 than the females) as well as ordinary variation. My own meas- 

 urements of adults show a still greater variation, the width of the 

 head entering the total length from 6.42 times to 11 times, al- 

 though allowance for a little exaggeration must here be made, in 

 that some of these measurements were from formalin specimens 

 while others were from chloretoned living animals. Formalin de- 

 creases the length out of proportion to the breadth. Still, as great 

 variation as that indicated by these figures undoubtedly exists. I 

 have made no experiments especially intended for the production 

 of extreme forms, and my records of extreme types have nearly 

 all been made during a single season only. I have preferred to 

 work for the causes of a variation that was perfectly obvious, 

 rather than merely to accumulate figures. It is, however, of inter- 

 est to point out that this degree of variation in general body pro- 

 portions is equal or in excess of that pointed out by Cope for the 

 entire genus Amblystoma, in which he recognized twelve Ameri- 

 can species. Complete measurements for all the species are not 

 given; but such measurements are given for nine species, show- 

 ing a range of variation from 8.06 to 10.79, ^.nd such partial 

 measurements as are given for the remaining species indicate that 

 their complete measurements, if given, would extend these limits' 

 little if at all. 



Now this great variation, which Cope assigns to irregularities 

 in metamorphosis or to the age of the adults, I find to be largely 



203 



