To these arguments we may begin by replying that the larva’s change of diet and the substitution of one morsel for a number are diametrically opposed to what happens before our eyes. The Hunting Wasp, as we know her, is extremely loyal to old customs; she has sumptuary laws which she never transgresses. She who fed on Weevils in her youth puts Weevils and naught else in her larva’s cell; she who was supplied with Buprestis-beetles persists in the fare which she has adopted and serves her larva with Buprestis-beetles. One Sphex must have Crickets; a second, Grasshoppers; a third, Locusts. Nothing is accepted but these particular dishes. The Bembex who hunts Gad-flies revels in them and refuses to do without them, whereas Stizus ruficornis, who fills the larder with Praying Mantes, scorns any other game. And so with the rest. They have each their own taste. It is true that many allow themselves a more varied bill of fare, but only within the limits of one entomological group: thus the Weevil and Buprestis hunters prey upon any species proportioned to their strength. Were the Hairy Ammophila to make a change in her diet, that would be her case too. Whether small and sundry to each cell or large and single, the prey would always consist of caterpillars. So far, so good. But there remains the question of the many replaced by the unit; and I do not yet know one instance of such an alteration in the Wasp’s habits. She who stocks the burrow with a single joint never thinks of heaping up several of smaller size; she who goes on repeated expeditions to stack a quantity of game in the same cell does not know how to limit herself to one head by choosing larger meat. The result of my observations never varies in this regard. The prehistoric Ammophila, who abandoned her multiplicity of small game for one colossal head, has nothing to warrant her existence. If the point were conceded, would the question be advanced? Not in the least. Let us accept as the initial prey a feeble caterpillar, paralysed with a single sting. Even then that sting must not be given at random, else the act would be more harmful than profitable. Irritated, but not subdued by the wound, the animal would but become more dangerous. The dart must strike a nerve-centre, probably in the middle region of the string of ganglia. This, at any rate, is how the present-day Ammophilæ seem to go to work when they are addicted to the rape of frail and slender grubs. What chance would the operator have of striking that one particular point, if her lancet were wielded without method? The probability is ludicrously remote: it is as one to the countless number of points whereof the caterpillar’s body is made up. And yet, according to the theorists, it is on this probability that the Wasp’s future depends. What an edifice to balance on the point of a needle!