others getting it not at all or only in small quantities. The sensitiveness of the larva to differences in nutrition must, however, be more finely graduated in termites than in bees; for Grassi has shown that even the distinction between workers and soldiers is under the control of the food providers. If the soldiers are taken away from them, they make new ones from other larvae.
NOTE XV (p. 44).
Herbert Spencer has attempted to weaken my reference to the social insects as an illustration of the metamorphosis of an organism taking place apart from any transmission of acquired characters (Cf. Weismann, 'The All-Sufficiency of Natural Selection; a Reply to Herbert Spencer,' Contemp. Rev., Sept. and Oct., 1893), by maintaining that the characters of the neuters had originated while all the females were still prolific. I will not deny the possibility of some of the peculiarities of the workers having actually originated as far back as this, though it would be difficult or even impossible to be at all certain on this point. But it seems to me that there is no room for doubt that the greater part of the characteristics of the workers have arisen later—that is, during and after the formation of castes. Herbert Spencer looks upon the infertility as a direct consequence of inadequate nourishment of the larvae. I have just shown that this cannot be so; but if, for a moment, we assume the contrary, how could we account for the instinct of the workers at the present day, which leads them to take far less food than the fertile females when in the imago-stage? Surely this cannot be explained by the meagreness of the nourishment during the larval stage! In my experiments with flies, at any rate, those which were poorly fed showed no less appetite as imagines than their sisters which had been well fed as larvae. Or ought we to suppose that the habit dates from ancient times, when all the females instinctively took but little food? In this case they would
