These weights speak eloquently for the vigorous huntress, and I never wearied of watching how swiftly and easily she resumed her flight, and rose out of sight with the game between her feet when approached too closely. But she did not always fly away, and then, though it was difficult to do so, and yet avoid hurting her, I would make her drop the prey by worrying and upsetting her with a straw. Then I would take possession of the victim, and the Cerceris, thus despoiled, would hunt about, go into her hole for a moment, come out, and resume the chase. In less than ten minutes the sharp-sighted insect would find a new victim, murder it and carry it off, not seldom to my profit. Eight times running have I stolen from the same individual; eight times did the indefatigable Cerceris resume her fruitless journey. Her perseverance tired out mine, and I let her keep the ninth capture. By this means, and by breaking open cells already filled with provisions, I got nearly a hundred weevils, and in spite of what I had a right to expect from what Léon Dufour has told us of the habits of the Cerceris bupresticida, I could not repress my astonishment at the sight of the singular collection which I had made. His Cerceris, though it limits itself to one genus, yet takes any species within that limit, but the more exclusive C. tuberculata preys exclusively on Cleonus ophthalmicus. On looking through my booty I met with but one single exception, and that belonged to a closely allied species, C. alternans—one which I never met with again in my frequent visits to the Cerceris. Later researches furnished me with a second exception, Bothynoderes albidus, and these are all. Can a specially succulent and savoury prey explain this predilection for a single species? Do the larvæ find in this unvaried diet juices which suit them peculiarly, and which they would not find elsewhere? I do not think so, and if Léon Dufour’s Cerceris hunted all the kinds of Buprestids, no doubt it was because they all have the same nutritive properties. But this must generally be the case with all the Curculionidæ; their alimentary properties must be identical, and in that case this amazing choice can only be one of size, and therefore of economy of labour and time. Our Cerceris, the giant of its race, chooses C. ophthalmicus as the largest in our district, and perhaps the commonest. But if this favourite prey fail, it must fall back upon other species, even if smaller, as is proved by the two exceptions above mentioned.