374 ON CHANGE OF MEXICAN AXOLOTL TO AN AMBLYSTOMA. 



presence hy tliefact that it may occasionally appear as a reversionary form. 

 Thus the Triton, auder some circumstances, sinks back to the Perenui- 

 branchiate stage, but not in such a way that the individual would first 

 become a Triton and then be transformed back to a Perennibranchi- 

 ate, but, as I have already made prominent above, simply by no longer 

 reaching the stage of the Salamandrida and remaining upon the stage 

 of the Ichthyodea. Thus, also, according to my hypothesis, the Sala- 

 mandrida that formerly lived on the shores of the Lake of Mexico, the 

 Amhlystoma Mexicauum, has sunk back to the stage of the fishsalaman- 

 der, and the only trace which remains to us of his former height of de- 

 velopment is just the inclination, more or less present in every individ- 

 ual, to reach under favorable circumstances the salamander stage again. 



But the third and last consequence which my explanation of the facts 

 brings with it lies in the altered part which would be assigned by it to 

 reversion in organic nature. Hitherto atavistic forms have been re- 

 garded only as isolated, exceptional cases, interesting, to be sure, in a 

 high degree for our knowledge, but without significance for the course 

 of development of organic nature. Kow a real importance would have 

 to be allowed them in this latter regard. 



I should assume that reversion may in a twofold manner be a con- 

 trolling power for the preservation or restoration of a form of life. In 

 one case, as in the Axolotl, where the newer form, standing organically 

 higher, becomes untenable from external causes, and now, as a further 

 development in the other direction does not seem i^ossible, instead of 

 simply dying out, a reversion of the species to the older and less highly 

 organized step follows. But, second, in this manner, that the older phy- 

 leticform is not altogether given tij), while the younger is developed from 

 it, hut that it alternates periodically tcith the younger, as we see in the 

 season-dimorphous butterflies. One will hardly urge any objection to 

 it if I regard the alternation of summer and winter form in these as a 

 periodically occurring reversion to the jihyletically older form (the winter 

 form). 



Though the total reversion of a species, as I assume it for the Axolotl, 

 may be a rarely-occurring case, the jyeriodically or cyclically occurring 

 reversion surely is not; it certainly plays a considerable part in the 

 origin of various forms of the alternating or cyclical mode of reproduc- 

 tion. 



POSTSCRIPT. 



It was intimated in the foregoing discussion that the causes from 

 which I derived the reversion of the hypothetical Amhlystoma Mexi- 

 canuni to the Axolotl of today, did not seem to me to suffice completely 

 for the explanation of the phenomenon. For one thing, they appeared 

 to me of too local a nature, as they could only be applied with certainty 

 to the Axolotl from the lake of the Mexican capital, while also the Paris 

 Axolotl, coming from another part of Mexico, requires an explanation 

 that will apply to him. But, on the other hand, they did not seem to 



