The cells with abundant provisions, reduced to a half or a third by my intervention, contain cocoons as small as the male cocoons, pale, translucent and limp, whereas the normal cocoons are dark-brown, opaque and firm to the touch. These, we perceive at once, are the work of starved, anaemic weavers, who, failing to satisfy their appetite and having eaten the last grain of pollen, have, before dying, done their best with their poor little drop of silk. Those cocoons which correspond with the smallest allowance of food contain only a dead and shrivelled larva; others, in whose case the provisions were less markedly decreased, contain females in the adult form, but of very diminutive size, comparable with that of the males, or even smaller. As for the controls which I was careful to leave, they confirm the fact that I had males in the part near the orifice of the reed and females in the part near the knot closing the channel. Is this enough to dispose of the very improbable supposition that the determination of the sex depends on the quantity of food? Strictly speaking, there is still one door open to doubt. It may be said that experiment, with its artifices, does not succeed in realizing the delicate natural conditions. To make short work of all objections, I cannot do better than have recourse to facts in which the experimenter's hand has not intervened. The parasites will supply us with these facts; they will show us how alien the quantity and even the quality of the food are from either specific or sexual characters. The subject of enquiry thus becomes double, instead of single as it was when I plundered one cell in my split reeds to enrich another. Let us follow this double current for a little while. An Ammophila, the Silky Ammophila (Cf. "The Hunting Wasps": chapter 13.--Translator's Note.), which feeds on Looper caterpillars (Known also as Measuring-worms, Inchworms, Spanworms and Surveyors: the caterpillars of the Geometrid Moths.--Translator's Note.), has just been reared in my refectory on Spiders. Replete to the regulation point, it spins its cocoon. What will emerge from this? If the reader expects to see any modifications, caused by a diet which the species, left to itself, had never effected, let him be undeceived and that quickly. The Ammophila fed on Spiders is precisely the same as the Ammophila fed on caterpillars, just as man fed on rice is the same as man fed on wheat. In vain I pass my lens over the product of my art: I cannot distinguish it from the natural product; and I defy the most meticulous entomologist to perceive any difference between the two. It is the same with my other boarders who have had their diet altered.