This state of partial immersion continues for half an hour or more. During this time, and even when the insect is roaming up and down the grass or twig, the posterior segments of the abdomen are extended at intervals, the abdomen turning upward at the same time. It is a kind of reaching-up movement, but whether this action accompanies a discharge of fluid or is an attempt at reaching for air I have not ascertained. Suddenly the insect begins to make bubbles by turning its tail out of the fluid, opening the posterior segment, which appears like claspers, and grasping a moiety of air, then turning the tail down into the fluid and instantly allowing the inclosed air to escape (Fig. 3). These movements go on at the rate of seventy or eighty times a minute. At the outset the tail is moved alternately to the right and left in perfect rhythm, so that the bubbles are distributed on both sides of the body, and these are crowded toward the head till the entire fluid is filled with bubbles, and the froth thus made runs over the back and around the stem (Fig. 4). Even when partially buried in these bubbles the tail is oscillated to the right and left, though when completely immersed the tail is only occasionally thrust out for air which is allowed to escape in the mass apparently without the right-and-left movement, though of this I am not sure. It is interesting to observe that in half a minute some thirty or forty bubbles are made in this way--a bulk of air two or three times exceeding the bulk of the body--without the slightest diminution in the size of the body.