574 HYMENOPTEKA The larvae are of two sorts. Those of species with a petiolated abdomen are white, footless grubs, incapable of any extended motion ; nor is this necessary, for they have not to seek their food, which is provided for them by their mothers, either collected and stored up for them in nests made by the female, as with bees and wasps, or by the eggs being placed in the bodies of other insects on which they live as parasites, or in galls upheaved on plants. It is, however, of interest to note that, while the larva after leaving the egg shows no trace of legs, yet they were present when it was in the egg. This shows clearly that the legs were lost through disuse. And this view is confirmed by the fact that the larvae of the lowest division of the Hymenoptera, those with a sessile abdomen, have jointed thoracic legs, and often abdominal legs as well, while they, for the most part, lead a free existence. In the lowest group of this division, the Siricidcc, as well as in those Tenthredinidce most nearly allied to them, the larvae have only the thoracic or true legs. They live either boring in wood, as Sirex, or in stems of plants, as Cephus, or in leaves rolled together by silken threads, as Lyda. Except Lyda, Cephus, and Xyela, all the saw-fly larvae have the usual thoracic and a variable number of false or abdominal legs, which are in fact merely prolongations of the ventral surface of the body, as with many phytophagous Colcoptera ; nor have they any thing like the claspers of lepidopterous caterpillars. With their similar habits and the presence of these ventral legs, saw-fly larvae have a considerable resemblance to the caterpillars of Lepidoptera, but they may be known from them in two ways, by always having more than five pairs of ventral legs (eight in Cimbcx, seven in Ne matus, and six in some Hylotomides), and never having more than two ocelli one on each side of the head instead of six on each side as in Lepidoptera. As with lepidopterous caterpillars, green is the prevailing colour of saw-fly larvae. They agree with them, too, in their general habits : they live on the leaves of various plants, devouring them in different ways, roll down leaves, raise galls, and mine leaves. It is interesting to note further that, as with lepi dopterous caterpillars, larvoa which are innoxious and eaten by birds are either entirely green, or green with black, pink, or white stripes along the sides and back, while noxious larvae those with bad smells or secretions which render them unpalatable to birds have bright contrasting colours with irregular markings, tubercles, &c., and they feed exposed, so that they may be readily seen and avoided. Some hymenopterous larvae before becoming pupae moult a certain number of times. According to Packard, a Bombus casts off its skin ten times ; Tenthredinidce do it five times ; but many (all parasitic, and most aculeates) do not moult until they become pupce; nor do they empty the- contents of the stomach till then. The period during which a hymenopteron remains in the larva state is seldom long ; it may be eight or nine days, or a month or two, but this depends on the season of the year; for many larvae, which have not been able to reach maturity in the autumn, remain in the same condition until the following spring, when they pupate. In order that this period may be passed in quietness, a cocoon is usually spun by the larva. For this purpose it is provided with a spinning apparatus, and a gland for secreting the silk required for the con struction of the cocoon. With the Aculeata it is thin and almost transparent ; it is of a firmer consistency with ichneumons, and is often coloured black, brown, or grey. Microgaster and other Bra- conidce spin their cocoons in company, and often around the dead body of the larva which they have devoured. They may be placed together without any regularity, or closely pressed in regularly ar ranged rows like the cells in a hive. Some ichneumons suspend their cocoons from twigs, &c. , by means of a silken thread. Gene rally the cocoon is single, but certain saw-flies (e.g., Cimbex) spin double ones, a thin inner one (which may be separated from the outer one by a considerable space) placed inside an outer, harder, and more tenacious covering. A cocoon, however, is not always spun. Cynipidce never spin one, nor apparently do Chalcididce nor some ants, e.g., Myrmica. The empty skin of the caterpillar which they have devoured is utilized by some ichneumons instead of a co coon ; galls serve the same purpose with others ; while Emphytus and other Tenthredinidce bore into pithy stems, where they pass into pupse without any other protection. Finally, others make a cell in the earth for the same purpose. The pupa resembles very closely the perfect insect, save that the wings are not developed, although visible as pad-like structures along the sides. The legs and antennae are laid along the front of the body, enclosed in thin pellicles. In certain Chalcididce (Eulo- phus] the pupa is of that form called " coarceate " ; that is to say, the entire body is enveloped in a case which conceals its form, and this case is hard and of some thickness, instead of being a mere thin transparent skin. When the insect leaves this pupal covering the latter retains its form intact. Most pupae are white. Many saw-fly pupse are green, with orange spots on the abdomen. A few Braconidce (Apanteles) have orange-coloured pupae. In the pupa state Hymenoptera remain but a short time, as a rule, not more than seven to ten days. They are not entirely quies cent in this condition, but have considerable power of motion, especially if exposed to the light, or disturbed in any way. Astho insect gradually reaches maturity, the pupa becomes more or less black, at first on the back of the thorax, then on the abdomen and limbs. When maturity is gained, the insect splits the pellicles which so closely envelop its body ; the limbs are freed ; the wings spread out and lose their flabby consistency ; the nervures become hard and firm ; the insect moves about, ejects from the anus a coloured liquid, and enters on its new mode of existence. Hymenopterous insects procreate by the union of the The two sexes. This takes place usually in the sunshine, sexes. Sometimes the connexion does not last more than a few seconds, and is not preceded by any preliminary courting. On the other hand many bees remain united for hours, and the genital parts of the male get torn and ruptured, so that it dies immediately after. Some species of Ckalcididce have been observed coquetting together for more than an hour before uniting. Generally copulation takes place on the ground ; but a few forms pair on the wing. The male Anttiophora, for instance, carries the female with him into the air for the marriage flight. This is the reason why the male Anthophora is larger than the female, instead of being smaller, as is usually the case. While, as has been said, Hymenoptera reproduce by the union of the two sexes, yet parthenogenesis or virgin repro duction is of not uncommon occurrence, and has been observed in all the families whose development for more than one generation can be traced with sufficient facility and accuracy. We meet with this phenomenon under (broadly speaking) two or three phases. Many females, if they cannot get access to males, will readily lay eggs, which are fertile and give issue to larvae ; but these larvae, when they reach maturity, yield invariably males. Any one can test this for himself with the too common gooseberry grub, Nematus ribesii. The same thing occurs with wasps, ants, and bees. With saw-flies, again, there are some species whose males are quite unknown, although the species have been caught and bred from the larvae in hundreds, e.g., Eriocampa ovata. Others have males, but they are ex tremely rare, e.g., Nematus gallicola. It has been shown, too, that the species just mentioned and some others (Phyl- lotoma nemorata, Poecilosoma pulveratum, &c.) with males unknown readily lay fertile eggs. The queen bee can lay eggs which will produce males or females, by opening or closing the spermatic sac, and letting the eggs come in con tact or not with the spermatic fluid. In the former case females will be the result, in the latter males. Worker bees, wasps, and ants deposit eggs which produce, however, only males. A still more curious phenomenon in connexion with the reproduction of some gall -insects (Cynipidce) requires to be mentioned. In early spring will be found on oak leaves and flowers soft, juicy, greenish, globular pea-shaped galls. Out of these come in summer the gall flies represented by both sexes. In the autumn (also on oak leaves) are found those curious flat brownish galls commonly called " oak spangles," which by many are taken for fungi, and have indeed been described as such. These " spangle galls " retain very much the same form during the autumn and winter ; then in March they swell up and become juicy, and a larva makes its appearance ; this soon becomes a pupa and finally a fly, but only in one sex, the female. It was long supposed that these two insects had no relationship with each other, that they belonged in fact to two distinct genera, for not only did the galls differ, but the insects themselves differed in the form of the body, the wings, &c. But it has recently been shown by Dr Adler that the two are forms of the same species, that there is an alternation of a spring bisexual form, with an autumnal unisexual one. A dimorphism of another kind exists among the social bees, wasps, and ants. An ordinary colony of these insects consists of three sorts of individuals. There is
the large female which founded the colony; then there
