munity, these parts would decrease in size; if the females no longer assist in building, if their entire duty is to lay eggs, their wings, legs, jaws, etc., will decrease; the surplus of force thus entailed being added to the reproductive system. Thus, then, there would have been produced the four castes found among the termites; the soldier representing the typical male of the species, the workers the typical females minus the perfected reproductive organs. But—and here is the great difficulty—how can the changed male have given his lost organs to his sex, and the female have transmitted her original but now modified peculiarities to the workers; especially as neuters do not propagate, and hence can not transmit their characters to progeny? Another law solves a portion of the difficulty: peculiarities acquired at any period of life are apt to appear in the same sex at the same time of life. Disuse having wrought its changes on our fertile insects after they had reached their perfect form, we can not expect them to appear in offspring which never reach this form. The modifications of structure were produced that the reproductive system might be benefited; why, then, should they take place in those insects which have a rudimentary and hence useless reproductive system?
We have now to ask whether those larvæ which are to produce fertile insects resemble, in any stage of their existence, the larvæ which are to produce neuters; whether, for example, the fertile male termite resembles, at any time, the infertile soldier. The reply to this is partly positive, partly negative; the larva of the female termite resembles very nearly the larva of the worker; but there is no great resemblance between the male and the soldier larvæ; there is a greater resemblance between the pupæ.
The fact alone that the female of social bees, and the male and female of white ants, should be presented under two forms is no novelty in insect history. In the aphides or plant-lice a similar state of affairs obtains: there is the perfect and imperfect female. Even as high in the scale as butterflies dimorphism is not uncommon. Mr. Wallace has discovered two forms of the female of Papilio Memnon, an inhabitant of the East Indies, one of which has tailed wings, the other of which is tailless. Several butterflies have three kinds of females, or are trimorphic. Hence the mere differences between neuter and perfect insects are nothing unusual, considered as differences. Were the workers fertile and thus able to propagate their peculiarities, the difficulties would vanish; but the problem why a fertile female should give birth to two or three distinct forms is still shrouded by mysteries, accept what explanation we may.
Our alternative that the workers are the type from which the males and the females have diverged is only an hypothesis, but in view of the facts it is the only alternative left us; since the neuters themselves can not have diverged from any type on account of their sterility.
