The larva does not readily change into a pupa in confinement. If, however, we wish to rear a beetle, the larva must be plentifully supplied with food, and removed from a bottle to a flat dish, where earth can be placed against the side so as to rise above the water-level. Fig. 77, where a pupa is represented in a bank by the side of a pond, will give us a hint how to go to work. The time occupied in pupation will vary according to the temperature of the room--but is never less than a fortnight. In the open it is probable that the winter is passed in the pupal condition, the perfect insect emerging in the spring. Like Land Beetles, it does not assume its dark hue for some days, but its pale skin darkens by degrees. [Illustration: FIG. 77.--Pupa of Dytiscus.] The larva of Hydrophilus does not seem to be often taken in this country. It would make a capital subject for investigation, and the food-supply could be arranged easily enough. The repetition of the experiences of Lyonnet, who reared these larvae from the egg, would be of great interest. He says that he took about thirty larvae from the brood, and fed them with very small water-snails. These they devoured in the same way as the larger larvae do. Having seized the snail with their mandibles, they bent backwards, and supporting it on their back, which served them for a table, eat it there, without making any use of their legs to hold their prey. When the supply of small snails ran short, they did very well with large ones cut up into pieces, and with tadpoles. If, however, food was not supplied to them, they fed on each other. But, except when pressed by hunger, they lived together peaceably enough, and seemed to take pleasure in each other’s society[60]. The larva of _Limnobia replicata_, a Crane-fly allied to, but smaller than, the well-known Daddy-longlegs, is another excellent subject for investigation. It is not very often taken; perhaps because it is not often looked for. But it is common enough, for all that. In general appearance these larvae resemble small caterpillars covered with spines. Some of these are simple and others forked, not much unlike the letter [image], with a short stem, and the arms slightly curved. There are no feet, and the last segment carries two pairs of hooks, one large and the other small. From their position they are called anal hooks. The dykes of the Sussex marshes are an excellent hunting-ground. Probably the channels of water-meadows, or any other shallow standing water in which aquatic moss grows, would serve the collector’s purpose quite as well. And such pieces of water abound all over the country. For taking these larvae the ordinary net and bottle are of little use, though a few may be captured by sweeping backwards and forwards among patches of aquatic moss.