Fabre took a partly grown Bembex larva from the nest, where it was surrounded by the remains of twenty flies. He fed it generously, and it ate sixty–two more, making a total of eighty–two in the eight days that passed before the spinning of the cocoon. Our experiments in this line gave similar results. We took charge of a partly grown larva on the afternoon of August tenth, and between that date and August fifteenth, when it spun its cocoon, it ate forty–two house flies besides a big Tabanus. Fabre thinks that under natural conditions the mother does not give the larva all it can eat at one time, but provides it with what she considers a reasonable amount of food, and keeps anything that she catches beyond this out of its reach. He draws his conclusion from the fact that he has found several flies in the tunnel leading to the nest, while the larva had as many more close to it. It would certainly be convenient for Bembex to have a reserve of this kind in case of rainy weather, but the forethought implied in such an action seems to require a higher degree of intelligence than can be claimed for her. In one nest we found a single fly with a long cylindrical egg attached to the left side of the thorax just at the origin of the third leg. In another, which we had seen made and provisioned, we found, six days later, a larva which we judged to be four days old. Assuming that the egg was laid on the first day, it must have taken it about two days to hatch. Other nests gave us larvæ in all stages of development, surrounded by the remains of Diptera, among which Syrphus, Tabanus, and Musca were represented. In regard to the condition of the flies captured by Bembex, we have never seen the crushing of the thorax, which is noted by both Wesenberg and Fabre. Indeed, the flies that we found were not always dead, since in two instances they responded readily to stimulation. Similar results have been obtained by Mr. S. W. Dunning of Hartford, Connecticut. Twice we have seen our spinolæ, as she was bringing home her prey, alight near the nest and sting it as it was held with the second pair of legs. We could see the process distinctly, since she is slow and clumsy, and, in one instance, had difficulty in reaching the fly, falling over to one side in an awkward manner. It is probable, then, that this is a habit with the wasp, but that the sting is usually given at the place of capture. We opened a number of Bembex nests, but succeeded in raising only one larva, which we took when it was half grown. This one, during the five days that passed before it spun the cocoon, ate forty–three flies. Mr. Bates has some notes on Monedula signata, which takes nothing but flies, and even confines itself to a single species, although it must sometimes go half a mile away to find it. This reminds us of Pompilus quinquenotatus, which never takes anything but Epeira strix. [Illustration: BEMBEX]