with sexual reproduction ; some of these must be more 
adapted to the needs of the animals than others. Those 
best adapted must be preserved, while the rest are 
destroyed. Let us grant the first two conditions to be 
true to nature. What chance is there that the requisite 
selection will be made ? Those which were most fitted to 
breathe the air of the atmosphere would either rise to do 
so as frequently, more frequently, or less frequently than 
those who were less endowed. If all rose to the surface 
to breathe the atmosphere as frequently, all would be 
equally exposed to danger, but the fish-eating enemy 
would make no choice, he would seize any individual 
which happened to come in his way. If the better en 
dowed individuals rose more frequently, the danger would 
be greater than in the case of the less endowed. If the 
better endowed rose less frequently, not needing to come 
up to the surface so often, and thus enjoyed greater 
security, the others, by using their lungs more frequently, 
would soon obliterate the difference brought about by the 
slight advantage of birth variation. 
For the reasons just assigned, I believe that the change 
of the tadpole into the frog was originally brought about, 
in the far-off ancestral history of the race, by the trans 
forming influence of outward circumstances, and not by 
Natural Selection. And this explanation avails to solve 
another problem. There are certain animals which were 
once regarded as belonging to entirely distinct species, 
such, for example, as the axolotl and the amblystoma. 
Certainly they differ from one another in a very marked 
degree. And yet observation and experiment have proved 
that the axolotl under certain conditions can be trans 
formed into the amblystoma. 
“ In 1867 some axolotls were observed to emerge from the water in 
the Jardin des Plantes at Paris, to cast their skins, and to become