There is, however, a far easier and surer method. A good handful of the moss should be gathered, and put into a shallow vessel half full of water—a developing-dish answers capitally—and then shaken to and fro or gently stirred with a small piece of stick. The larvae will curl into a ring and fall to the bottom, whence they may be picked up and dropped into a bottle or other receptacle to be taken home. A fair quantity of moss should be gathered, for this is their favourite food, and all larvae are greedy feeders.

Other water-plants, however, do not come amiss to them. Just before these lines were written I was examining a bottle in which some of these larvae were kept. It contained a few sprays of willow-moss and some ivy-leaved duckweed, which floated on the surface. One larva on a spray of moss was reaching upward, and it was distinctly seen to feed on the duckweed. This must have been from choice, for there was within reach plenty of what all observers consider to be its natural food. This, too, might have been more easily obtained; for, to reach the duckweed, the larva had to hold on to the moss by the anal hooks, and extend its body in a fashion analogous to that of the caterpillar of a geometer moth, which will hold on to a branch with its fore-legs and claspers and maintain the body at an angle of 45°, sometimes for hours together. I have also seen them feed on hornwort.

The larva of this small Crane-fly is not at all difficult to keep. It is exceedingly hardy, and will survive a great deal of rough treatment. In November, 1895, I sent three in a tube—securely packed, as I thought—to a friend in Yorkshire. A few days afterwards I heard, with regret, ‘that the bottle was broken in transit, and that the larvae arrived dead.’ Three days later I was gratified by receiving a letter, from which the following is quoted: ‘The Limnobia larvae have come to life again. I put them into water as soon as they arrived, and after lying motionless for many hours they have begun to creep about and feed.’

This is excellent testimony to their powers of endurance, but it is weak in comparison to that which De Geer supplies[61].

He was Marshal of the Court of Sweden, and one November, before leaving his country house for his official duties at Stockholm, he put four of these larvae into a vessel of water, and left them to take their chance. The water froze into a solid mass. When he returned in the following May he found about half the water thawed, and two of the larvae dead. The others, though they had been enclosed in ice all the winter, were lively and in good condition. He put them into another vessel with fresh water and some aquatic moss, and at once they began to move about and commenced to feed. Both pupated by the fifteenth of the month, and the perfect fly emerged after six days in the pupal stage.

Fig. 78.—Larva of Limnobia replicata.

The following description of this larva is principally condensed from De Geer, whence the figures are also taken. The body is long and cylindrical, and divided into eleven segments, of which the first and largest is sub-triangular. The second and third segments are shorter than the rest.

The head (t) is very small and completely retractile within the first segment, the anterior margin of which completely closes the orifice, so that, in this condition, the creature appears to be headless. The body is covered with spines, some simple and others branched. On the first three segments there are only simple spines; but from the fourth to the tenth segment inclusive there are also on each segment two forked spines—that is, fourteen in all. On the last segment the spines are simple, and here are also four chitinous hooks, one pair much larger than the other. These simple spines carry a white vessel, which extends throughout their whole length; but in the forked spines there are two such vessels placed side by side in the stem, and diverging one to each branch.

He comes to the conclusion that these spines are probably the breathing apparatus of the larva, for they are similar to those which he observed in an aquatic caterpillar. Walker[62] calls these spines ‘long filamentous processes, which appear to be internally supplied with air-tubes,’ but he does not seem to have taken the trouble to break up a specimen, or he would have been in no doubt as to their real nature. This is shown by the fact that the larva never comes to the surface to take in a supply of air, but contents itself with the oxygen dissolved in the water.