Second act of this wondrous spectacle. In order to witness it, we must open a window in cell upon cell until fortune deigns to smile upon us. The larva is hatched and already fairly large. Like the egg, it hangs perpendicularly, by the rear, from the ceiling; but the suspensory cord has gained considerably in length and consists of the original thread eked out by a sort of ribbon. The grub is at dinner: head downwards, it is digging into the limp belly of one of the caterpillars. I touch up the game that is still intact with a straw. The caterpillars grow restless. The grub forthwith retires from the fray. And how? Marvel is added to marvels: what I took for a flat cord, for a ribbon, at the lower end of the suspensory thread, is a sheath, a scabbard, a sort of ascending gallery wherein the larva crawls backwards and makes its way up. The cast shell of the egg, retaining its cylindrical form and perhaps lengthened by a special operation on the part of the new-born grub, forms this safety-channel. At the least sign of danger in the heap of caterpillars, the larva retreats into its sheath and climbs back to the ceiling, where the swarming rabble cannot reach it. When peace is restored, it slides down its case and returns to table, with its head over the viands and its rear upturned and ready to withdraw in case of need. Third and last act. Strength has come; the larva is brawny enough not to dread the movements of the caterpillars' bodies. Besides, the caterpillars, mortified by fasting and weakened by a prolonged torpor, become more and more incapable of defence. The perils of the tender babe are succeeded by the security of the lusty stripling; and the grub, henceforth scorning its sheathed lift, lets itself drop upon the game that remains. And thus the banquet ends in normal fashion. That is what I saw in the nests of both species of the Eumenes and that is what I showed to friends who were even more surprised than I by these ingenious tactics. The egg hanging from the ceiling, at a distance from the provisions, has naught to fear from the caterpillars, which flounder about below. The new-hatched larva, whose suspensory cord is lengthened by the sheath of the egg, reaches the game and takes a first cautious bite at it. If there be danger, it climbs back to the ceiling by retreating inside the scabbard. This explains the failure of my earlier attempts. Not knowing of the safety-thread, so slender and so easily broken, I gathered at one time the egg, at another the young larva, after my inroads at the top had caused them to fall into the middle of the live victuals. Neither of them was able to thrive when brought into direct contact with the dangerous game. If any one of my readers, to whom I appealed just now, has thought out something better than the Eumenes' invention, I beg that he will let me know: there is a curious parallel to be drawn between the inspirations of reason and the inspirations of instinct. CHAPTER 12. THE OSMIAE.