336 SEXUAL SELECTION. [Part IL 



kingdom, the greater size of the females seems generally 

 to depend on their developing an enormous number of ova ; 

 and this may to a certain extent hold good with insects. 

 But Dr. Wallace has suggested a much more probable ex- 

 planation. He finds, after carefully attending to the de- 

 velopment of the caterpillars of Bomhyx cynthia and 

 yama-mal, and especially of some dwarfed caterpillars 

 reared from a second brood on unnatural food, " that in 

 proportion as the individual moth is finer, so is the time 

 required for its metamorphosis longer ; and for this reason 

 the female, which is the larger and heavier insect, from 

 having to carry her numerous eggs, will be preceded by 

 the male, which is smaller and has less to mature." " Now, 

 as most insects are short-lived, and as they are exposed to 

 many dangers, it would manifestly be advantageous to the 

 female to be impregnated as soon as possible. This end 

 would be gained by the males being first matured in large 

 numbers ready for the advent of the females; and this 

 again would naturally follow, as Mr. A. R. Wallace has 

 remarked," through natural selection ; for the smaller 

 males would be first matured, and thus would procreate a 

 large number of offspring which would inherit the reduced 

 size of their male parents, while the larger males from be- 

 ing matured later would leave fewer offspring. 



There are, howcAcr, exceptions to the rule of male in- 

 sects being smaller than the females ; and some of these 

 exceptions are intelligible. Size and strength would be 

 an advantage to the males, which fight for the possession 

 of the female; and m these cases the males, as with the 

 stag-beetle (Lucanus), are larger than the females. There 

 are, liowever, other beetles which are not known to fight 

 together, of which the males exceed the females in size ; 

 and the meaning of this fact is not known ; but in some 



" ' Transact. Ent. Soc' 3d series, vol. v. p. 486. 

 1* ' Journal of Proc. Ent. Soc' Feb. 4, 186Y, p. Ixxi. 



