200 EXPERIMENTAL EVOLUTION LECT. 



more suitable than vegetable food. He has shown 

 also that growth is certainly correlated with other 

 influences, since, even when food is superabundant, 

 tadpoles become sooner transformed into the adult 

 form when they are few than when they are numerous 

 in the same space. This had already been noticed by 

 Semper in his experiments on Lymnaea stagnates* to 

 which reference has already been made, and I have 

 also spoken of my own experiments on the same 

 subject. 2 To these facts I shall merely add the 

 general fact, recorded by Herbert Spencer, that trout 

 and other fishes living in small streams are generally 

 small ; that pond-snails grow larger in ponds than in 

 rivers, and that under natural conditions, as Balas- 

 chewa has noticed, the smaller the mass of water in 

 which molluscs live, the smaller they remain. And 

 R. P. Whitfield, 3 who has performed experiments 

 similar to those of Semper, has obtained the same 

 results, with this additional interesting circumstance, 

 that when the dwarfing process is made to operate 



1 C. Semper, Ueber die Wachsthums-Bedingungen der Lymnaeus 

 stagnalis. Arbeit en aus dem Zool. Zoot. Institut zu Wiirzburg, vol. i., 

 1874. 



2 H. de Varigny, Contribution Experimental a f Etude de la Crois- 

 sance. Comptes Rendus, June 15, 1891. 



3 Description of the Animal of Lymnaea megasoma (Say), with 

 some Account oj the Changes produced by Confinement in Aquaria 

 and under Unnattiral Conditions. Am. Naturalist, 1880, p. 51 

 (abstract). 



