Thus in our second example, she stung the first three segments in the regular order, the third, the second, and lastly (and most persistently) the first. She then went on, without a pause, to sting the fourth, fifth, sixth, and seventh, stopping at this point and leaving the posterior segments untouched. In our first example, it will be remembered, the middle segments were spared. The stinging being completed, she proceeded to the process known as _malaxation_, which consists in repeatedly squeezing the neck of the caterpillar, or other victim, between the mandibles, the subject of the treatment being turned around and around so that all sides may be equally affected. [Illustration: AMMOPHILA URNARIA STINGING CATERPILLAR] In our third case a caterpillar which we had caught was placed in front of a wasp just after she had carried the second larva into her nest. She seemed rather indifferent to it, passing it once or twice as she ran about, but finally picked it up and gave it one prolonged sting between the third and fourth segments. She then spent a long time in squeezing the neck, pinching it again and again, after which it was left on the ground; and as she showed no further interest in it we carried it home for further study. In the three captures, then, that came under our observation, all the caterpillars being of the same species and almost exactly of the same size, three different methods were employed. In the first, seven stings were given at the extremities, the middle segments being left untouched, and no malaxation was practiced. In the second, seven stings again, but given in the anterior and middle segments, followed by slight malaxation. In the third, only one sting was given, but the malaxation was prolonged and severe. Let us now compare these variations with those of Fabre. In his first case the sting entered at twelve different points, beginning between the first and second segments and progressing regularly backward. There was no malaxation. In his second example the third, second, and first segments were stung in the order given, and thereafter each succeeding segment up to the ninth, nine stings being given in all, with careful malaxation following. In his later experiments, which seem to have been numerous, he found that as a usual thing all the segments were stung, although the posterior three or four were occasionally spared, but that the order in which they were operated upon, as well as the amount of malaxation, was very variable. Our conclusions, then, as to Ammophila’s methods of stinging agree fairly well with those of Fabre; but there is one important exception. In his cases the middle segments, upon one of which the egg is laid in our species as well as in his, were invariably stung, and this he considers a point of extreme importance. In one of our cases the middle segments were not touched. [Illustration: CATERPILLAR WITH EGG OF AMMOPHILA URNARIA]