For now, having finished thrusting with her sting, the Odynerus clasps the creature and begins slowly, with great bites of the mandibles, to munch the last three segments, on the dorsal surface. A manifest gluttony accompanies these bites; all the mouth-parts are brought into play, as though the insect were feasting on some exquisite dish. Meanwhile the grub, bitten to the quick, desperately works its short legs, whose activity is not at all diminished by the stings administered behind; it struggles violently, protesting with its head and mandibles. The other takes no notice and continues gnawing at the larva’s rump. This lasts for ten or fifteen minutes; then the bandit releases the sufferer and leaves it where it lies, without troubling about it any further, instead of carrying it with her as she would not fail to carry game intended for the nest. Soon afterwards, the Odynerus begins to lick her fingers, as though she had been consuming some toothsome dainty: time after time she passes her tarsi between her mandibles; she is washing her hands after rising from table. What has she been eating? I must once more watch the epicure squeeze the juice from the rump. Ever obliging, provided that I practise a little patience, my six captives, one after the other, operate on the Chrysomela-larvæ, at one time in front, as game for the family, at another behind, as a little addition to their own diet. The honey with which I serve them on spikes of lavender does not make them forget this horrible treat. The tactics employed in obtaining it, though the same in the general aspect, vary in detail. The larva is always seized by the hinder end and the stings are administered in succession from back to front, on the ventral surface. Sometimes the abdomen only is attacked, sometimes the thorax also, when the victim is deprived of all movement. Evidently the object of these stings is not the immobility of the larva, since the latter can move quite well, ambling along, wounded though it be, when the sting has not gone higher than the abdomen. Inertia is indispensable only in the case of victuals intended for the cells. If the Odynerus is working on her own behalf and not for her family, it matters little to her whether the grubs whose dainties she covets struggle or not; it is enough if all resistance in the part to be exploited is abolished by paralysis. This paralysis, moreover, is quite accessory; and each huntress neglects or practises it at will, bearing more or less forward, without any fixed rule. When the sated Odynerus releases the grub whose rump she has been chewing, it is sometimes therefore inert, like those intended for the cells, and sometimes endowed with almost as much activity as the untouched grubs, from which it differs only by the absence of its anal pimple, its support which reminds us of a cripple sitting in a bowl.