6l2 



SCIENCE PROGRESS 



it is also, I think, obvious that the classification adopted is 

 in broad outlines correct. 



The Axolotl is of peculiar interest. As is well known, this 

 animal is normally neotenous — i.e. never metamorphoses, but 

 becomes sexually mature in the larval form. Cases of spon- 

 taneous metamorphosis do, however, occur. It is not generally 

 known, however, that there are two geographical races of the 

 species which differ markedly on that point. The variety 

 from the non-mountainous part of the United States (New 

 Mexico) is very prone to spontaneous metamorphosis, so much 

 so indeed that Dr. Swingle writes to me that he considers it 

 almost useless for experimental work on metamorphosis. In the 

 specimens from the Rocky Mountains and the highlands of 



Fig. 3. 



F"*", F, F". Curves of production of female-determining substance in various 

 strains of Lymantria. 



M+, M, M". Corresponding curves for male-determining substance. 



Ti-T^. Points where sex-development is transformed from female to male; 

 occurring in crosses of F- ? x M+ c7, F" ? x M cJ and F ? x M+ <? respectively. 



Old Mexico, however, spontaneous metamorphosis is exceed- 

 ingly rare. That it does occur is well established ; indeed the 

 knowledge that the Axolotl is not a Perennibranchiate, but a 

 neotenous larva of a species of Amblystoma, was due to such 

 a case occurring under Dumeril's eyes in 1863 (8). But it has 

 always been very uncommon, and in the strains now bred by 

 aquarium dealers in Europe appears to be practically non- 

 existent. This fact is very probably correlated with the fact 

 that the old Mexican race inhabits large permanent lakes and 

 mountain streams ; while that from New Mexico Uves mainly 

 in small pools Uable to dry up in summer. Be that as it ma}^ 

 the mountain race is of great experimental importance, as it 

 provides an " all-or-nothing " metamorphic reaction. 



The Axolotl's thyroid is not markedly small for its body 

 size, and is still, as grafting experiments show, capable of 



