excessively minute for creatures of such complex organization. Hymenoptera are probably less widely distributed than Aptera, Coleoptera or Diptera, but they are to be found in all except the most inhospitable regions of the globe. The order is, with few exceptions, terrestrial or aerial in habit. Comparatively only a few species are, for part of their lives, denizens of fresh water; these, as larvae, are parasitic on the eggs or larvae of other aquatic insects, the little hymenopteron, Polynema natans, one of the “ fairy-fiies ”-swims through the water by strokes of her delicate wings in search of a dragon-fiy's egg in which to lay her own egg, while the rare Agriotypus dives after the case of a. caddis-worm. It is of interest that the waters have been invaded by the parasitic group of the Hymenoptera, since in number of species this is by far the largest of the order. No group of terrestrial insects escapes their attacks-even larvae boring in wood are detected by ichneumon flies with excessively long ovipositors. Not a few cases are known in which a parasitic larva is itself pierced by the ovipositor of a “ hyper parasite, ” and even the offspring of the latter may itself fall a victim to the attack of a “ tertiary parasite.”
Fossil History.-Very little is known of the history of the Hymenoptera previous to the Tertiary epoch, early in which, as we know from the evidence of many Oligocene and Miocene fossils, all the more important families had been differentiated. Fragments of wings from the Lias and Oolitic beds have been referred to ants and bees, but the true nature of these remains is doubtful. Classijication.-Linnaeus divided the Hymenoptera into two sections-the Terebrantia, whose females possess a cutting or piercing ovipositor, and the Aculeata, in which the female organ is modified into a sting. This nomenclature was adopted by P. A. Latreille and has been in general use until the present day. A closely similar division of the order results from T. Hartig's character drawn from the trochanter-whether of two segments or undivided-the groups being termed respectively Ditrocha and Monotrocha. But the most natural division is obtained by the separation of the saw-flies as a primitive sub-order, characterized by the imperfect union of the first abdominal segment with the thorax, and by the broad base of the abdomen, so that there is no median constriction or “ waist, ” and by the presence of thoracic legs-usually also of abdominal pro-legs-in the larva. All the other families of Hymenoptera, including the gall-flies, ichneumon's and aculeates, have the first abdominal segment closely united with the thorax, the second abdominal segment constricted so as to form a narrow stalk or “ waist, ” and legless larvae without a hinder outlet to the food-canal. These two suborders are usually known as the Sessiliventra and Petiolivenlra respectively, but the names Symlphyta and A pocrita proposed in 1867 by C. Gerstaecker have priority, and should not be replaced. Symphyta.
This sub-order, characterized by the “ sessile, " broad-based abdomen, whose first segment is imperfectly united with the thorax, and by the usually cater illar-like larvae with legs, includes the various groups of saw-flies. Three leading families may be mentioned. The Cephidaé, or stem saw-flies, have an elongate pronotum, a compressed abdomen, and a single spine on the shin of the fore-leg. The soft, white larvae have the thoracic legs very small and feed in the stems of various plants. Cephus pygmaeus is a well-known enemy of corn crops. The Siricidae (“ wood-wasps ) are large elongate insects also with one spine on each fore-shin, but with the pronotum closely joined to the mesothorax. The ovipositor is long and prominent, enabling the female insect to lay her eggs in the wood of trees, where the white larvae, whose legs are excessively short, tunnel and feed. These insects are adorned with bands of black and yellow, or with bright metallic colours, and on account of their large size and formidable ovipositors they often cause needless alarm to persons unfamiliar with their habits. The Tenlhredinidae, or true saw-Hies, are distinguished b two spines on each fore-shin, while the larvae are usually caterpillars, with three airs of thoracic legs, and from six to eight pairs of abdominal prolegs, the latter not posses sin the hooks found on the pro-legs of lepidopterous caterpillars. Nfost saw-fly larvae devour leaves, and the beautifully serrate processes of the ovipositor are well adapted for egg-laying in plant tissues. Some saw-Hy larvae are protected by a slimy secretion (fig. 6, c) and a few live concealed in galls. In the form of the feelers, the wing-neuration and minor structural details there is much diversity among the saw-flies. They have been usually regarflctl as a single family, but W. ll. Ashmcad has lately differentiated eleven families of them. A pacrita.
This sub-order includes the vast majority of the Hymenoptera, characterized by the narrowly constricted waist in the adult and by* the legless condition of the larva. The trochanter is simple in some genera and divided in others. With regard to the minor divisions of this group, great difference of opinion has revailed among students. In his recent classification Ashmead figol) recognizes seventy-nine families arranged under eight “ super-families.” The number of species included in this division is enormous, and the multiplication of families is, to some extent, a natural result -of increasingly close study. But the distinctions between many of these rest on comparatively slight characters, and it is likely that d Q 8
msg . .- 4' 1s
4 A. A Q~ ' g;;;“="i3' Q
I '€§ , .
L' *, ' fl/
TE RZI. . '-After
Marlatt, Ent. Circ. 26, U.S. Dept. Agric. FIG. 6.—a, Pear Saw-Hy (Eriocampoides limacina); b, larva without, and c, with its slimy protective coat; e, cocoon; f, larva before pupation; g, pupa, magnified; d, leaves with larvae. the future discovery of new genera may abolish many among such distinctions as may now be drawn. It seems advisable, therefore, in the present article to retain the wider conception of the family that has hitherto contented most writers on the Hymenoptera. Ashmead's “super-families” have, however, been adopted as founded on definite structural characters-they probably indicate relationship more nearly than the older divisions founded mostly on habit. The Cynipoidea include the gall-flies and their parasitic relations. In the Chalcidoiclea, lchneumonoidea and Proctotrypoidea will be found nearly all the “ parasitic Hymenoptera ” of older classifications. The Formicoidea are the ants. The group of Fossores, or “ digging-wasps, ” is divided by Ashmead, one section forming the Sphecoidea, while the other, together with the Chrysidae w< r .F »
3-'l'-¢'5 if ° i*
J/ /~ i l ' § °“
at b we 6,
After Howard, Ent Tech. Bull. 5 U.S. Dept. Agric. FIG. 7.-Chalcid (Dibffachys boucheanus), a hyper-parasite. a, Larva. b, Female fly.
d, Its head more highly magni- c, Pupa of male. fied. e, Feeler.
and the true wasps, make up the Vespoidea. The Apoidea consists of the bees only.
Cynipoidea.-In this division the ovipositor' issues fwm the ventral surface of the abdomen; the pronotum reaches back to the tegulae; the trochanter has two segments; the fore-wing (fig. 4, 2) has no stigma, but one or two areolets. The feelers with twelve to fifteen segments are thread-like and straight. All the insects included in this group are small and form two families-the Cynipidae and the Figitidae. They are the “ gall-flies, ” many of the species laying eggs in various plant-tissues where the presence of the larva causes the formation of a pathological growth or gall, always of a definite
form and characteristic of the species; the “ oak-apple ” and the