I, in my turn, remark similar facts in my two dozen grub-worms. They are curled in a ring; they are stacked one upon another, but with a certain confusion in the ranks; their backs touch the wall. I will not attribute this circular curve to the effect of the sting which was very probably administered, for I have never observed it in the caterpillars stabbed by the Ammophilæ; I believe rather that the position is natural to the grub during inaction, even as it is natural for the Iuli [25] to coil themselves into a spiral. In this living bracelet there is a tendency to return to the rectilinear conformation; it is a bent bow fighting against the obstacle that surrounds it. By the very fact, therefore, of being curled up, each grub keeps more or less steady by pressing its back a little against the wall; and it retains its place even when the cell approaches the vertical. Moreover, the shape of the cell has been calculated with a view to this manner of storing. In the part next the entrance, the part which one might call the store-room, the cell is cylindrical and narrow, so as to afford the living rings as little space as possible; they are thus kept in position and are unable to slip. It is here that the grubs are stacked, squeezed one against the other. At the other end, near the back, the cell expands into an ovoid to give the larva elbow-room. The differences between the two diameters is very perceptible. At the entrance I find only four millimetres: [26] at the back I find six. [27] Thanks to this inequality of width, the cell comprises two apartments: the provision-store in front and the dining-room behind. The Eumenes’ spacious cupola does not permit of this arrangement; there the game is heaped up in disorder, the oldest in date promiscuously with the most recent; and each piece is merely bent, not rolled. The ascending-sheath provides a remedy for the disadvantages of this confusion. Note also that the packing of the victuals is not the same from one end of the Odynerus’ skewerful to the other. In the cells whose provisions have not yet or have only recently been broached, I observe this detail: near the egg or the newly-hatched larva, in the part which I have just described as the dining-room, the space is not fully occupied; there are just a few grubs here, three or four, somewhat isolated from the bulk and leaving enough room to ensure the safety of either the egg or the young larva. This is the food supplied for the early meals. If there be danger in the first mouthfuls, which are the most risky of all, the life-line provides a means of withdrawal. More towards the front, the game is piled in close-packed layers, the stack of worms is continuous.