96 ' NATURAL SCIENCE. Feb.. 



more moderately put down the chances against a certain combination 

 as " milHons to one," and against a still more complex combination 

 as " billions to one." But a million millions are but ten multiplied 

 into itself eleven times. The selection or survival of one out of ten 

 during only a dozen consecutive generations, is equivalent to the 

 selection or survival of one out of a potential billion. The survival of one 

 out of two for forty generations gives a similar result. Allowing for 

 atavism and many other hindrances which will seriously retard the 

 rate of evolution of complex efficienc}', what objection can there be to 

 the supposition of slow complex evolution by survival of the fittest 

 daring, not a dozen or forty generations, but a thousand or a million ? 

 Mr. Spencer contrasts the facts that varieties of dogs cannot be 

 produced by differences of nutrition, and that differences of nutrition 

 are known to determine the caste in some of the social insects. He 

 holds that this proves that, in the case of social insects, " the produc- 

 tion of their various castes does not result from the Natural 

 Selection of varying germ-plasm." But what else could have 

 evolved the highly peculiar potentialities, or sets of potentialities, of a 

 germ-plasm capable of developing not merely into either of two sexes 

 (which is common to dogs and other animals), but into one, or two, 

 or three additional neuter castes besides ? The absence or non- 

 evolution of special susceptibilities to the influence of nutrition in 

 dogs is no disproof of their evolution by Natural Selection in bees 

 and ants. On the contrary, it seems to show that a long process of 

 evolution must have been required for their development in social 

 insects. No such additional complexities and susceptibilities of the 

 germ-plasm appear in mammals, where there have been no special 

 means or steps by which Natural Selection would be likely to 

 adequately favour and evolve them. They only appear in certain 

 egg-laying organisms which feed their helpless larvas to maturity, so 

 that the chances of the evolution of neuter helpers in the excep- 

 tionally arduous maternal task are of a highly favourable character. 

 Delay or loss of the reproductive portion of the maternal functions 

 in some of the offspring, if combined with early appearance of the 

 nursing instincts, would be an advantage in rearing a numerous 

 family of younger sisters and brothers, and this would greatly 

 promote the success of the queen and her fertile progeny. A dog 

 could not be thus readily or materially assisted by her pups either 

 in the growth of embryos within her, or in the suckling or feeding of 

 the young after birth. She does not rear completely helpless 

 offspring to complete maturity ; her family breaks up at an early 

 period. The evolution of mammals has been on different lines. 

 They feed their offspring for a time, indeed, but by no means to 

 maturity, as is necessary where the entire larval or caterpillar stage 

 has to be provided for until the full-grown imago or perfect insect 

 emerges. As neuter castes are only found where Natural Selection 

 has had special opportunities or facilities for such evolution, we may 



