Bramble-bees and Others 



to the point of disappearing altogether. We 

 see this especially in the third case, where the 

 presence of a solitary female in a family of 

 twenty-six is due to the somewhat larger dia- 

 meter of the corresponding Snail-shell and 

 also, no doubt, to some mistake on the 

 mother's part, for the female cocoon, in a 

 series of two, occupies the upper storey, 

 the one next to the orifice, an arrangement 

 which the Osmia appears to me to dislike. 



This result throws so much light on one 

 of the darkest corners of biology that I must 

 attempt to corroborate it by means of even 

 more conclusive experiments. I propose next 

 year to give the Osmiae nothing but Snail- 

 shells for a lodging, picked out one by one, 

 and rigorously to deprive the swarm of any 

 other retreat in which the laying could be ef- 

 fected. Under these conditions, I ought to 

 obtain nothing but males, or nearly, for the 

 whole swarm. 



There would still remain the inverse per- 

 mutation : to obtain only females and no 

 males, or vei*y few. The first permutation 

 .makes the second seem very probable, 

 although I cannot as yet conceive a means of 

 realizing it. The only condition which I can 



178 



