Swammerdam says of this larva, ‘When about to eat he seizes with the two teeth (mandibles) the little creatures that come in his way, and pierces their body with the sharp crooked points. The teeth being perforated from the points to the roots, he in a surprising manner sucks through them into his mouth the blood of the unfortunate captive. This may be easily seen, especially when the blood of his prey is of a red colour, as the teeth are transparent.’ He then describes how he watched one of these larvae feed, and saw the blood, mixed with air-bubbles, travel up the mandibles. After this he tells us how, if we have a Dytiscus larva, we may ‘procure ourselves a very entertaining and surprising sight, by throwing to it a small earthworm; for let this last move, twine, and otherwise bestir itself ever so much, the other keeps its hold, and very calmly sucks the blood of its prisoner.’ We may, however, go to work in another fashion. We may dissect out the mandibles from a dead larva and pass a fine hair into the slit near the point, and it will come out at the orifice near the base. It is through this orifice that the nutritive juices of the prey are drawn into the true mouth. This practical proof that the mandibles are pierced is by no means so difficult as one might suppose. De Geer[57] seems to have been the first to suggest that there must be some kind of true mouth, and in support of his suggestion tells us that he saw this larva eating up the solid parts of a Water Woodlouse (_cloporte_), after having sucked up its juices. More than this, he places the mouth in what has proved to be the true position, though he did not solve the mystery as to why it is kept so firmly closed. This was done by Mr. Burgess, an American naturalist, from whose paper[58] the following particulars are condensed:-- ‘Authors have described this creature as mouthless; and if we examine the slit where we should expect the mouth to be, we find that this slit ends in a perfectly closed seam. The methods of microscopical research were brought into play, and a longitudinal section of the head cut and mounted. This showed that the upper and lower lips were locked together by a peculiar joint--the upper coming over and locking into the under lip (Fig. 75).’ We may get some idea of this mouth-lock by placing the fingers of the right hand over those of the left, and then bending them. Mr. Burgess concludes his paper thus: ‘We find that the Water-tiger, far from being mouthless, as ordinarily assumed, has in fact a very wide mouth, though its lips are closely locked together by a dove-tailed grooved joint developed for this purpose. Whether this joint can be unlocked by the animal itself is another question, which I cannot answer, though De Geer’s observation above quoted makes this probable. It is, at all events, easy to open the mouth by manipulation with a pair of forceps.’ [Illustration: FIG. 75.--Mouth-lock. _m_, mouth × 125. (After Burgess.)]