But if the interest be great, the difficulty of observation is not small. The solitary habits of the Ammophila,—their being scattered singly over wide spaces, and their being almost always met with by mere chance,—almost forbid, as in the case of Sphex occitanica, any experiment being prepared beforehand. Long must a chance be watched for and awaited with unalterable patience, and one must know how instantly to profit by it when at last it comes just when least expected. I have waited for such a chance for years and years, and then, all at once, I got the opportunity with a facility for observation and clearness of detail which made up for the long waiting. At the beginning of my observations I succeeded twice in watching the murder of the caterpillar, and saw, as far as the rapidity of the operation allowed, that the sting of the Hymenopteron struck once for all at the fifth or sixth segment of the victim. To confirm this I bethought myself of making sure which ring was stabbed by examining caterpillars which I had not seen sacrificed, but had carried off from their captors while they were being dragged to the burrow; but it was vain to use a microscope,—no microscope can show any trace of such a wound. This was the plan adopted. The caterpillar being quite still, I tried each segment with the point of a fine needle, measuring the amount of sensibility by the greater or less pain given. Should the needle entirely transpierce the fifth segment or the sixth, there is no movement. But prick even slightly one in front or behind, the caterpillar struggles with a violence proportioned to the distance from the poisoned segment. Especially does the least touch on the hinder ones produce frantic contorsions. So there was but one stab, and it was given in the fifth or sixth segment. What special reason is there that one or other of these two should be the spot chosen by the assassin? None in their organisation, but their position is another thing. Omitting the Loopers of Ammophila holosericea, I find that the prey of the others has the following organisation, counting the head as the first segment:—Three pairs of true feet on rings two, three, and four; four pairs of membranous feet on rings seven, eight, nine, and ten, and a last similar pair set on the thirteenth and final ring; in all eight pairs of feet, the seven first making two marked groups—one of three, the other of four pairs. These two groups are divided by two segments without feet, which are the fifth and sixth.