A strange result! Honey, the nectar of the flowers, the sole diet of the apiary under its two forms and the sole nourishment of the predatory insect in its adult phase, is for the larva of the same insect an object of insurmountable disgust, and probably a poison. The transfiguration of the chrysalis surprises me less than this inversion of the appetite. What change occurs in the stomach of the insect that the adult should passionately seek that which the larva refuses under peril of death? It is no question of organic debility unable to support a diet too substantial, too hard, or too highly spiced. The grubs which consume the larva of the Cetoniae, for example (the Rose-chafers), those which feed upon the leathery cricket, and those whose diet is rich in nitrobenzine, must assuredly have complacent gullets and adaptable stomachs. Yet these robust eaters die of hunger or poison for no greater cause than a drop of syrup, the lightest diet imaginable, adapted to the weakness of extreme youth, and a delicacy to the adult! What a gulf of obscurity in the stomach of a miserable worm! These gastronomic experiments called for a counter-proof. The carnivorous grub is killed by honey. Is the honey-fed grub, inversely, killed by carnivorous diet? Here, again, we must make certain exceptions, observe a certain choice, as in the previous experiments. It would obviously be courting a flat refusal to offer a heap of young crickets to the larvae of the Anthophorus and the Osmia, for example; the honey-fed grub would not bite such food. It would be absolutely useless to make such an experiment. We must find the equivalent of the bee smeared with honey; that is, we must offer the larva its ordinary food with a mixture of animal matter added. I shall experiment with albumen, as provided by the egg of the hen; albumen being an isomer of fibrine, which is the principal element of all flesh diet. _Osmia tricornis_ will lend itself to my experiment better than any other insect on account of its dry honey, or bee-bread, which is largely formed of flowery pollen. I knead it with the albumen, graduating the dose of the latter so that its weight largely exceeds that of the bee-bread. Thus I obtain pastes of various degrees of consistency, but all firm enough to support the larva without danger of immersion. With too fluid a mixture there would be a danger of death by drowning. Finally, on each cake of albuminous paste I install a larva of medium growth.