Let us look into the burrow of a Wasp who stocks her grubs’ provisions in advance: if we select the moment when the insect is going in with its prey, we shall find in the cell a certain number of victims, the commencement of a larder, but never at that time a grub, nor even an egg, for this is not laid until the provisions are quite complete. When the egg is laid, the cell is closed and the mother does not return to it. It is therefore only in burrows where the mother’s visits are no longer necessary that we can find larvæ side by side with larger or smaller stocks of food. On the other hand, let us inspect the home of a Bembex at the moment when she is entering with the fruits of her hunting. We are certain of finding in the cell a larva, big or little as the case may be, among remnants of provisions already consumed. The portion which the mother is now bringing is therefore intended to prolong a meal which has already lasted several days and which is to continue for some time further with the produce of future hunting expeditions. Should we be fortunate enough to make this search towards the end of the larva’s infancy—an advantage which I have enjoyed as often as I wished to—we shall find, on a copious heap of remnants, a large and portly grub, to which the mother is still bringing fresh victuals. The Bembex does not cease her catering and does not leave the cell for good until the larva, distended by a purply paste, refuses its food and lies down, stuffed to repletion, on the jumble of legs and wings of the game which it has devoured. Each time that the mother enters the burrow on returning from the chase, she brings but a single Fly. If it were possible, by counting the remnants contained in a cell whose occupant is full-grown, to tell the number of victims supplied to the larva, we should know how often at the least the Wasp visited her burrow after laying the egg. Unfortunately, these broken victuals, chewed and chewed again at moments of scarcity, are for the most part unrecognizable. But, if we open a cell with a less forward nurseling, the provisions lend themselves to examination, some of them being still whole or nearly whole, while others, more numerous, are represented by fragments in a state of preservation that enables them to be identified. Incomplete though it be, the list obtained under these conditions is surprising and shows what activity the Wasp must display to satisfy the needs of such a table. I will set forth one of the bills of fare which I have observed.