Fig. 301—Larva of honey-bee: g, brain; bm, ventral nervous cord; œ, œsophagus; sd, spinning-gland; cd, mid-intestine or chyle-stomach; ed, hind-intestine, not yet connected with the mid-intestine; vm, urinary tube; an, anus; st, stigmata.—After Leuckart, from Lang.
In certain larval insects, as those of the Proctotrypidæ (first larval stage), the higher Hymenoptera (ichneumons, ants, wasps, and bees, Fig. 301), in the Campodea-like larvæ of the Meloidæ and Stylopidæ, the larva of the ant-lion (Myrmecoleo), and those of Diptera pupipara (Melophagus), the embryonic condition of the separation of the proctodæum and mid-gut (mesenteron) persists, the stomach ending in a blind sac; in such cases the intestine, together with the urinary tubes, is entirely secretory.
The anus is wanting in the larva of the ant-lion, as also in the wasps (in which there is a rudimentary colon) and in freshly hatched bees, though it becomes perfectly formed in the fully grown larvæ (Newport, art. Insecta, p. 967, and H. Müller).
In the larvæ of lamellicorn Coleoptera (Melolontha vulgaris) the digestive tube is nearly as simple as in bees, though there is a large colon, which at its beginning forms an immense cœcum, and has also one anal aperture (Newport).
The length and shape of the digestive canal is dependent on the nature of the food and also on the mode of life, especially the ease or difficulty with which the food is digested.
Newport, while stating that the length of the alimentary canal in larvæ is not in general indicatory of the habits of the species, makes this qualification after describing the digestive canal of Calandra as compared with that of Calosoma: “The length and complication of the intestines, therefore, appear to have some reference to the quality of the food to be digested, since it is well known that the food of these latter insects (weevils) is of difficult assimilation, being as it is chiefly the hard ligneous fibres of vegetable matter; but they cannot be received as always indicatory of a carnivorous [or] vegetable feeder, since, as above remarked, the length of the canal is considerable in one entirely carnivorous larva, while it is much shorter in some herbivorous, and particularly in pollenivorous larvæ, as in the Melolontha and the apodal Hymenoptera.”
Fig. 302.—Digestive canal of a carabid beetle: b, œsophagus; c, crop; d, proventriculus; f, mid-intestine, or “chyle-stomach,” with its cœeca; g, posterior division of the stomach; i, the two pairs of urinary tubes; h, intestine; k, rectum; l, anal glands.—After Dufour, from Judeich and Nitsche.
Newport also contends that the length of the alimentary canal is not more indicative in the perfect insect of the carnivorous or phytophagous habits of the species than in the larva. It is nearly as long (being from two to three times the length of the whole body), and is more complicated, in the rapacious Carabidæ (Fig. 302) than in the honey-sipping Lepidoptera, whose food is entirely liquid. Referring to the digestive canal of Cicindelidæ, which is scarcely longer than the body, he claims that “we cannot admit that the length of the digestive organs, and the existence of a gizzard and gastric vessels, are indicatory of predacity of habits in the insect, because a similar conformation of parts exists often in strictly vegetable feeders. The existence and length of these parts seem rather to refer to the comparative digestibility of the food than to its animal or vegetable nature.” Newport then refers to the digestive canal of Forficulidæ (in which the gizzard is present, the canal, however, passing in an almost direct line through the body, making but one slight convolution), “a farther proof that the length of the canal must not be taken as a criterion whereby to judge of the habits of a species.” He adds this will apply equally well to the omnivorous Gryllidæ, in which there exists a short alimentary canal, but a gizzard of more complicated structure than that of the Dytiscidæ.