Polygamy by massacre of males, as among herbivora, and gallinaceæ seems a step toward a more logical and more economic distribution of the sexes. If antelopes perpetuate themselves very well with one male to an hundred females, is it not an indication that a part at least of the sacrificed males might have dispensed with being born? And would it not be better, in the interest of the antelopes, that a part of these males, if they ought to continue to be bora, should be normally sexless, as with termites, and entrusted with some social duty? The organization of termites is very pretty; it will do to finish off this brief review of animal societies founded on the unsexing of sexes. One has already noted, in the chapters on dimorphism, the diversity of sexual forms, corresponding to four quite distinct castes. The minute examination of one of their republics permits one to assert differentiations much more numerous, for each of the principal castes passes through active larval and nymphal forms, adolescent forms, such as most neuroptera and libellules also present. In taking count of all the nuances one may observe in a state (to use the familiar word) of termites fifteen different forms, all with marked characteristics. The principal are: 1. Workers, 2. Soldiers, 3. Small males, 4. Small females, 5. Large males, 6. Large females, 7. Nymphs with little cases, 8. Nymphs with long cases, 9. Larvæ. When one attacks an ant hill, the soldiers arrive at the breach, very threatening, odd, with their bodies all head, all mandibles. The enemy routed, the workers come to repair the damage. There are sometimes several female egg-layers; sometimes there is only one male: copulation always takes place outside the hill, and as with ants, the males perish, while the fecundated females become the origin of a new state. The expeditions of travelling termites, common as fighting termites in South Africa, are naturally directed by soldiers. Sparmann (cited in Guérin's Dictionnaire d'histoire naturelle) observed them during his voyage to the Cape, and says they behave rather as non-coms in close rank, or climbing onto grass blades, watch the defile, beating with their feet, if the order were bad, or too slow. The signal is at once understood, and obeyed by the rank at once, is answered by a whistle. There is in this something so marvellous that one hesitates to accept the traveller's interpretation in entirety. It is not the spontaneous and mechanical discipline of the ants, but the consenting obedience, so difficult to obtain from inferior humanities. After all, nothing is impossible, and without being credulous in these matters, one need be astonished at nothing. Nevroptera are, moreover, exceeding old on the earth; they date from before the coal-beds: their civilization is some thousands of centuries older than human civilizations.