He describes the filaments on the body of the larva, and concluded that they were probably breathing-organs, because of the dark-coloured vessels within them. These he traced, as we will presently do, to their union in the stem of the gill, and thence to the air-vessels running down each side of the body of the larva. He fed them on leaves of the Water Soldier, and kept them through the winter. In the following June he found them preparing to undergo their transformation into the pupal stage, and at the end of the month the moths came out. He was gratified by seeing the congress of these insects. The females deposited their eggs on the floating vegetation and on the sides of his aquarium, a little below the surface of the water, and in about eight days the young larvae were seen. [Illustration: FIG. 82.--Larva of _Paraponyx stratiotata_ (enlarged).] These larvae must be very abundant, though they do not seem to be often taken by collectors. In describing an allied (American) species, which is found on Vallisneria and pondweed, Mr. Hart says[65], ‘They feed at first exposed on the leaf, but later two or even three leaves are loosely webbed together, face to face, by each larva, between which it remains concealed while feeding. They are, therefore, hard to discover unless their hiding-places are broken up by seining, or the like, when the larvae may be seen swimming about.’ This is, no doubt, the reason why these larvae are not more often taken. Anything like a seine net is of course out of the question for us; but masses of vegetation may be readily broken up by vigorously working the bottle and net backwards and forwards amongst them. Specimens I have seen were taken among duckweed; and Mr. Hart mentions one instance of part of the larval case being constructed of ivy-leaved duckweed, ‘which was abundant there at that time.’ Now let us bring our lens to bear, so that we may make out the external structure, and recognize the similarity of the breathing-organs of this Moth-larva to those of the Crane-fly larva already treated of (p. 168).