1911 Encyclopædia Britannica/Hexapoda

by David Sharp and George Herbert Carpenter

HEXAPODA (Gr. ἕξ, six, and πούς, foot), a term used in systematic zoology for that class of the Arthropoda, popularly known as insects. Linnaeus in his Systema naturae (1735) grouped under the class Insecta all segmented animals with firm exoskeleton and jointed limbs—that is to say, the insects, centipedes, millipedes, crustaceans, spiders, scorpions and their allies. This assemblage is now generally regarded as a great division (phylum or sub-phylum) of the animal kingdom and known by K. T. E. von Siebold’s (1848) name of Arthropoda. For the class of the true insects included in this phylum, Linnaeus’s old term Insecta, first used in a restricted sense by M. J. Brisson (1756), is still adopted by many zoologists, while others prefer the name Hexapoda, first used systematically in its modern sense by P. A. Latreille in 1825 (Familles naturelles du règne animal), since it has the advantage of expressing, in a single word, an important characteristic of the group. The terms “Hexapoda” and “hexapod” had already been used by F. Willughby, J. Ray and others in the late 17th century to include the active larvae of beetles, as well as bugs, lice, fleas and other insects with undeveloped wings.

Characters.

A true insect, or member of the class Hexapoda, may be known by the grouping of its body-segments in three distinct regions—a head, a thorax and an abdomen—each of which consists of a definite number of segments. In the terminology proposed by E. R. Lankester the arrangement is “nomomeristic” and “nomotagmic.” The head of an insect carries usually four pairs of conspicuous appendages—feelers, mandibles and two pairs of maxillae, so that the presence of four primitive somites is immediately evident. The compound eyes of insects resemble so closely the similar organs in Crustaceans that there can hardly be reasonable doubt of their homology, and the primitively appendicular nature of the eyes in the latter class suggests that in the Hexapoda also they represent the appendages of an anterior (protocerebral) segment. Behind the antennal (or deutocerebral) segment an “intercalary” or tritocerebral segment has been demonstrated by W. M. Wheeler (1893) and others in various insect embryos, while in the lowest insect order—the Aptera—a pair of minute jaws—the maxillulae—in close association with the tongue are present, as has been shown by H. J. Hansen (1893) and J. W. Folsom (1900). Distinct vestiges of the maxillulae exist also in the earwigs and booklice, according to G. Enderlein and C. Börner (1904), and they are very evident in larval may-flies. The number of limb-bearing somites in the insectan head is thus seen to be seven. All of these are to be regarded as primitively post-oral, but in the course of development the mouth moves back to the mandibular segment, so that the first three somites—ocular, antennal and intercalary—lie in front of it. In Lankester’s terminology, therefore, the head of an insect is “triprosthomerous.” The maxillae of the hinder pair become more or less fused together to form a “lower lip” or labium, and the segment of these appendages is, in some insects, only imperfectly united with the head-capsule.

The thorax is composed of three segments; each bears a pair of jointed legs, and in the vast majority of insects the two hindmost bear each a pair of wings. From these three pairs of thoracic legs comes the name—Hexapoda—which distinguishes the class. And the wings, though not always present, are highly characteristic of the Hexapoda, since no other group of the Arthropoda has acquired the power of flight. In the more generalized insects the abdomen evidently consists of ten segments, the hindmost of which often carries a pair of tail-feelers, (cerci or cercopods) and a terminal anal segment. In some cases, however, it can be shown that the cerci really belong to an eleventh abdominal segment which usually becomes fused with the tenth. With very few exceptions the abdomen is without locomotor limbs. Paired processes on the eighth and ninth abdominal segments may be specialized as external organs of reproduction, but these are probably not appendages. The female genital opening usually lies in front of the eighth abdominal segment, the male duct opens on the ninth.

In all main points of their internal structure the Hexapoda agree with other Arthropoda. Specially characteristic of the class, however, is the presence of a complex system of air-tubes (tracheae) for respiration, usually opening to the exterior by a series of paired spiracles on certain of the body segments. The possession of a variable number of excretory tubes (Malpighian tubes), which are developed as outgrowths of the hind-gut and pour their excretion into the intestine, is also a distinctive character of the Hexapoda.

The wings of insects are, in all cases, developed after hatching, the younger stages being wingless, and often unlike the parent in other respects. In such cases the development of wings and the attainment of the adult form depend upon a more or less profound transformation or metamorphosis.

With this brief summary of the essential characters of the Hexapoda, we may pass to a more detailed account of their structure.

Exoskeleton

The outer cellular layer (ectoderm or “hypodermis”) of insects as of other Arthropods, secretes a chitinous cuticle which has to be periodically shed and renewed during the growth of the animal. The regions of this cuticle have a markedly segmental arrangement, and the definite hardened pieces (sclerites) of the exoskeleton are in close contact with one another along linear sutures, or are united by regions of the cuticle which are less chitinous and more membranous, so as to permit freedom of movement.

Head.—The head-capsule of an insect (figs. 1, 2) is composed of a number of sclerites firmly sutured together, so that the primitive segmentation is masked. Above is the crown (vertex or epicranium), on which or on the “front” may be seated three simple eyes (ocelli). Below this comes the front, and then the face or clypeus, to which a very distinct upper lip (labrum) is usually jointed. Behind the labrum arises a process—the epipharynx—which in some blood-sucking insects becomes a formidable piercing-organ. On either side a variable amount of convex area is occupied by the compound eye; in many insects of acute sense and accurate flight these eyes are very large and sub-globular, almost meeting on the middle line of the head. Below each eye is a cheek area (gena), often divided into an anterior and a posterior part, while a distinct chin-sclerite (gula) is often developed behind the mouth.

From Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 1.—Head and Jaws of Cockroach (Blatta). Magnified 10 times. A, Front; B, side; C, back; v, vertex; f, frons; cl, clypeus; lbr, labrum; oc, compound eye; ge, gena; mn, mandible; ca, st, pa, ga, la, cardo, stipes, palp, galea, lacinia of first maxilla; sm, m, pa′, pg, sub-mentum, mentum, palp, galea of 2nd maxilla.

Feelers.—Most conspicuous among the appendages of the head are the feelers or antennae, which correspond to the anterior feelers (antennules) of Crustacea. In their simpler condition they are long and many-jointed, the segments bearing numerous olfactory and tactile nerve-endings. Elaboration in the form of the feelers, often a secondary sexual character in male insects, may result from a distal broadening of the segments, so that the appendage becomes serrate, or from the development of processes bearing sensory organs, so that the structure is pinnate or feather-like. On the other hand, the number of segments may be reduced, certain of them often becoming highly modified in form.

After Marlatt, Entom. Bull. 14, n. s. (U.S. Dept. Agric.).

Fig. 2.—Head of Cicad, front view. Ia, frons;
b, clypeus (the pointed labrum beneath it); II, mandible; III, first maxilla; (a, base; b, sheath;
c, piercer), III′, inner view of sheath; IV, second maxillae forming rostrum (b, mentum; c, ligula).

Jaws.—The mandibles of the Hexapoda are usually strong jaws with one or more teeth at the apex (fig. 1, A, B, mn), articulating at their bases with the head-capsule by sub-globular condyles, and provided with abductor and adductor muscles by means of which they can be separated or drawn together so as to bite solid food, or seize objects which have to be carried about. They never bear segmented limbs (palps) and only exceptionally (as in the chafers) is the skeleton composed of more than one sclerite. The mandibles often furnish a good example of “secondary sexual characters,” being more strongly developed in the male than in the female of the same species. In most insects that feed by suction the mandibles are modified. In bugs (Heteroptera) and many flies, for example, they are changed into needle-like piercers (fig. 2, II), while in moths and caddis-flies they are reduced to mere vestiges or altogether suppressed.

As previously mentioned, a pair of minute jaws—the maxillulae—are present in the lowest order of insects, between the mandibles and the first maxillae. They usually consist of an inner and an outer lobe arising from a basal piece, which bears also in some genera a small palp (see Aptera).

In their typical state of development, the first maxillae offer a striking contrast to the mandibles, being composed of a two-segmented basal piece (cardo and stipes, fig. 1, C, ca, st) bearing a distinct inner and outer lobe (lacinia and galea, fig. 1, C, la, ga) and externally a jointed limb or palp (fig. 1, C, pa). Such maxillae are found in most biting insects. In insects whose mouths are adapted for sucking and piercing, remarkable modifications may occur. In many blood-sucking flies, for example, the galea is absent, while the lacinia becomes a strong knife-like piercer and the palp is well developed. In bugs and aphids the lacinia is a slender needle-like piercer (fig. 2, III), while the palp is wanting. In butterflies and moths the lacinia is absent while the galea becomes a flexible process, grooved on its inner face, so as to make with its fellow a hollow sucking-trunk, and the palp is usually very small.

The second pair of maxillae are more or less completely fused together to form what is known as the labium or “lower lip.” In generalized biting insects, such as cockroaches and locusts (Orthoptera), the parts of a typical maxilla can be easily recognized in the labium. The fused cardines form a broad basal plate (sub-mentum) and the stipites a smaller plate (mentum)—see fig. 1, C, sm, m—jointed on to the sub-mentum, while the galeae, laciniae and palps remain distinct. In specialized biting insects, such as beetles (Coleoptera), the labium tends to become a hard transverse plate bearing the pair of palps, a median structure—known as the ligula—formed of the conjoined laciniae, and a pair of small rounded processes—the reduced galeae—often called the “paraglossae,” a term better avoided since it has been applied also to the maxillulae of Aptera, entirely different structures. The long sucking “tongue” of bees is probably a modification of the ligula. In bugs and aphids (Hemiptera), the fused second maxillae form a jointed grooved beak or rostrum (fig. 2, IV) in which the slender piercers (mandibles and first maxillae) work to and fro.

This second pair of maxillae (or labium) form then the hinder or lower boundary of the mouth. In front or above the mouth is bounded by the labrum, while the mandibles and first maxillae lie on either side of it. A median process, known as the hypopharynx or tongue, arises from the floor of the mouth in front of the labium, and becomes most variously developed or specialized in different insects. The salivary duct opens on its hinder surface. It does not appear to represent a pair of appendages, but the maxillulae of the Aptera become closely associated with it. According to the view of R. Heymons, the hypopharynx represents the sterna of all the jaw-bearing somites, but other students consider that it belongs to the mandibular and first maxillary segments, or entirely to the segment of the first maxillae.

Neck.—The head is usually connected with the thorax by a distinct membranous neck, strengthened in the more generalized orders with small chitinous plates (cervical sclerites). These have been interpreted as indicating one or more primitive segments between the head and thorax. Probably, however, as suggested by T. H. Huxley (Anat. Invert. Animals, 1877), they really belong to the labial segment which has not become completely fused with the head-capsule. It has been shown by C. Janet (1889), from careful studies of the musculature, that the greater part of the head-capsule is built up of the four anterior head-segments, the hindmost of which has the mandibles for its appendages, and this conclusion is in the main supported by the recent work on the head skeleton of J. H. Comstock and C. Kochi (1902) and W. A. Riley (1904).

Thorax.—The three segments which make up the thorax or fore-trunk are known as the prothorax, mesothorax and metathorax (see fig. 3). The dorsal area of the prothorax is occupied by a single sclerite, the pronotum (fig. 3, d), which is large and conspicuous in those insects, such as cockroaches, bugs (Heteroptera) and beetles, which have the prothorax free—i.e. readily movable on the segment (mesothorax) immediately behind—smaller and of less importance where the prothorax is fixed to the mesothorax, as in bees and flies. The dorsal area of the mesothorax, and also of the metathorax, may be made up of a series of sclerites arranged one behind the other—prescutum, scutum, scutellum and post-scutellum (fig. 3, e, f, g, h), the scutellum of the mesothorax being often especially conspicuous. Ventrally, each segment of the thorax has a sternum with which a median pre-sternum and paired episterna and epimera are often associated (see figs. 3, 4). The recent suggestion of K. W. Verhoeff (1904) that the hexapodan thorax in reality contains six primitive segments is entirely without embryological support.

Legs.—Each segment of the thorax carries a pair of legs. In most insects the leg is built up of nine segments: (1) a broad triangular, sub-globular, conical or cylindrical haunch (coxa); (2) a small trochanter; (3) an elongate stout thigh (femur); (4) a more slender shin (tibia); and (5-9) a foot consisting of five tarsal segments. The fifth (distal) tarsal segment carries a median adhesive pad—the pulvillus—on either side of which is a claw. The pulvillus is probably to be regarded as a true terminal (tenth) segment of the leg, while the claws are highly modified bristles. Numerous bristles are usually present on the thighs, shins and feet of insects, some of them so delicate as to be termed “hairs,” others so stout and hard that they are named “spines” or “spurs.” In the relative development and shape of the various segments of the leg there is almost endless variety, dependent on the order to which the insect belongs, and the special function—walking, running, climbing, digging or swimming—for which the limb is adapted. The walking of insects has been carefully studied by V. Graber (1877) and J. Demoor (1890), who find that the legs are usually moved in two sets of three, the first and third legs of one side moving with the second leg of the other. One tripod thus affords a firm base of support while the legs of the other tripod are brought forward to their new positions.

After Marlatt, Ent. Bull. 3, n. s. (U.S. Dept. Agr.).

Fig. 3.—Thorax of Saw-Fly (Pachynematus).

I, Dorsal view. II, Ventral view. III, Lateral view. IV, Lateral view with segments separated.

  Prothorax:

a, Episternum. b, Sternum. c, Coxa of fore-leg. d, Pronotum.

  Mesothorax:

e, Prescutum. f, Scutum. g, Scutellum. h, Post-scutellum. i, Mesophragma. j, Epimeron. k, Episternum. l, Coxa of middle leg.

  Metathorax:

m, Scutum. o, Epimeron. p, Coxa of hind leg. n, First Abdominal Segment. t, Tegula at base of fore-wing.

After Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 4.—Legs and Ventral Thoracic Sclerites of Female Cockroach (Blatta).

I, Fore-leg and pro-sternum (S) in front of which are the ventral cervical sclerites (c).
  cx, Coxa. tr, Trochanter. fe, Thigh. tb, Shin. ta, Tarsal segments.

II, Middle leg and mesosternum.

III, Hind-leg and metasternum. In IIIa, the episternum (a) and epimeron (b) are
  slightly separated.

Wings.—Two pairs of wings are present in the vast majority of insects, borne respectively on the mesothorax and metathorax. At the base of the wing, i.e. its attachment to the trunk, we find a highly complex series of small sclerites adapted for the varied movements necessary for flight. Those of the dragon-flies (Odonata) have been described in detail by R. von Lendenfeld (1881). The long axis of the wings, when at rest, lies parallel to the body axis. In this position the outer margin of the wing is the costa, the inner the dorsum, and the hind-margin the termen. The angle between the costa and termen is the apex. When the wing is spread, its long axis is more or less at a right angle to the body axis. A wing is an outgrowth from the dorsal and pleural regions of the thoracic segment that bears it, and microscopic examination shows it to consist of a double layer of cuticularized skin, the two layers being in contact except where they are thickened and folded to form the firm tubular nervures, which serve as a supporting framework for the wing membrane, enclose air-tubes, and convey blood. These nervures consist of a series of trunks radiating from the wing-base and usually branching as they approach the wing-margins, the branches being often connected by short transverse nervures, so that the wing-area is marked off into a number of “cells” or areolets.

After Quail, Natural Science, vol. xiii., J. M. Dent & Co.

Fig. 5.—Wing-Neuration in a Cossid Moth. 2, sub-costal; 3, radial; 4, median; 5, cubital; 6, 7, 8, anal nervures.

The details of the nervuration vary greatly in the different orders, but J. H. Comstock and J. G. Needham have lately (1898–1899) shown that a common arrangement underlies all, six series of longitudinal or radiating nervures being present in the typical wing (see fig. 5). Along the costa runs a costal nervure. This is followed by a sub-costal which sometimes shows two main branches. Then comes the radial—usually the most important nervure of the wing—typically with five branches, and the median with four. These sets arise from a main trunk towards the front region of the wing-base. From another hinder trunk arise the two-branched cubital nervure and three separate anal nervures. In the hind-wing of many insects the number of radial branches becomes reduced, while the anal area is especially well developed and undergoes a fan-like folding when the wings are closed. Great diversity exists in the texture and functions of fore and hind-wings in different insects; these differences are discussed in the descriptions of the various orders. The wings often afford secondary sexual characters, being not infrequently absent or reduced in the female when well developed in the male (see fig. 6). Rarely the male is the wingless sex.

In addition to the wings there are smaller dorsal outgrowths of the thorax in many insects. Paired erectile plates (patagia) are borne on the prothorax in moths, while in moths, sawflies, wasps, bees and other insects there are small plates (tegulae)—see Fig. 3, t—on the mesothorax at the base of the fore-wings.

Abdomen.—In the abdominal exoskeleton the segmental structure is very clearly marked, a series of sclerites—dorsal terga and abdominal sterna—being connected by pale, feebly chitinized cuticle, so that considerable freedom of movement between the segments is possible. The first and second abdominal sterna are often suppressed or reduced, on account of the strong development of the hind-legs. In many insects ten, and in a few eleven, abdominal segments can be clearly distinguished in addition to a small terminal anal segment. The female genital opening usually lies between the seventh and eighth segments, the male on the ninth. Prominent paired limbs are often borne on the tenth segment, the elongate tail-feelers (cerci) of bristle-tails and may-flies, or the forceps of earwigs, for example. In the Embiidae, a family of Isoptera, it has been shown by G. Enderlein (1901) that these cerci clearly belong to a partially suppressed eleventh segment, and R. Heymons (1895–1896) has proved by embryological study that in all cases they really belong to this eleventh segment, which in the course of development becomes fused with the tenth. Smaller appendages (such as the stylets of male cockroaches) may be carried on the ninth segment. Pairs of processes carried on the eighth and ninth segments often become specialized to form the ovipositor of the female (see fig. 14) and the genital armature of the male. A marked modification of the hinder abdominal segments may be noticed in most insects, the sclerites of the eighth and ninth being frequently hidden by those of the seventh. In the higher orders several of the hinder segments may be altogether suppressed.

From Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 6.—Outline of Male (♂) and Female (♀) Cockroaches (Blatta) from the side, showing Abdominal Segments (numbered 1-10).

Internal Organs

From Miall and Denny (after Newton), The Cockroach, Lovell Reeve & Co.

Fig. 7.—Brain of Cockroach from side. oe, Gullet; op, optic nerve; sb, sub-oesophageal ganglion; mn, mx, mx′, nerves to jaws; t, tentorium.

Nervous System.—The nervous system in the Hexapoda is built up on the typical arthropodan plan of a double ventral nerve-cord with a pair of ganglia in each segment, the cords passing on either side of the gullet and connecting with an anterior nerve-centre or brain (fig. 7) in the head. The brain innervates the eyes and feelers, and must be regarded as a “syncerebrum” representing the ganglia of the three foremost limb-bearing somites united with the primitive cephalic lobes. Behind the gullet lies the sub-oesophageal nerve-centre (fig. 7, sb), composed of the ganglia of the four hinder head-somites and sending nerves to the jaws. A pair of ganglia in each thoracic segment is usual (fig. 8), and as many as eight distinct pairs of abdominal ganglia may often be distinguished, the hindmost of which represents the fused ganglia of the last four segments. But in many highly organized insects a remarkable concentration of the trunk-ganglia takes place, all the nerve-centres of the thorax and abdomen in the chafers and in the Hemiptera, for instance, being represented by a single mass situated in the thorax. The legs, wings and other organs of the trunk receive their nerves from the thoracic and abdominal ganglia, and the fusion of several pairs of these ganglia may be regarded as corresponding to a centralization of individuality. A special “sympathetic” system arises by paired nerves from the oesophageal connectives; these nerves unite, and send back a median recurrent nerve associated with ganglia on the gullet and crop, whence proceed cords to various parts of the digestive system.

In connexion with the central nervous system there are usually numerous organs of special sense. Most insects possess a pair of compound eyes, and many have, in addition, three simple eyes or ocelli on the vertex. The nature of these organs is described in the article Arthropoda. The surface of a compound eye is seen to be covered with a large number of hexagonal corneal facets, each of which overlies an ommatidium or series of cell elements (fig. 9, A, B). There are over 25,000 ommatidia in the eye of a hawk moth.

After Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 8.—Ventral Muscles and Nerve Cord of Cockroach.

Auditory organs of a simple type are present in most insects. These consist of fine rods suspended between two points of the cuticle, and connected with nerve-fibres; they are known as chordotonal organs. In many cases a more complex ear is developed, which may be situated in strangely diverse regions of the insect’s body. In locusts (Acridiidae) a large ovate, tympanic membrane (fig. 9, G) is conspicuous on either side of the first abdominal segment; on the inner surface of this membrane are two horn-like processes in contact with a delicate sac containing fluid, connected with which are the actual nerve-endings. In the nearly-related crickets and long-horned grasshoppers (Locustidae) the ears are situated in the shins of the fore-legs (see fig. 9, F). Just below the knee-joint there is a swelling, along which two narrow slits run lengthwise. They lead into chambers, formed by inpushing of the cuticle, whose delicate inner walls are in contact with air-tubes; on the outer surface of these latter are ridges, along which the special nerve-endings are arranged. An ear of another type is found in the swollen second segment of the feeler in many male gnats and midges, the cuticle between this segment and the third forming an annular drum which is connected with numerous nerve-endings, while the fine bristles on the more distal segments vibrate in response to the note produced by the humming of the female.

From Ridley, Insect Life, vol. 7 (U.S. Dept. Agr.).

Fig. 9.—Single Ommatidium of Cockroach’s Eye (after Grenacher). B, Section through compound eye (after Miall and Denny); C, organs of smell in cockchafer (after Kraepelin); D, a, b, sensory pits on cercopods of golden-eye fly; c, sensory pit on palp of stone-fly (after Packard); E, sensory hair (after Miall and Denny); F, ear of long-horned grasshopper; a, Front shin showing outer opening and air-tube; b, section (after Graber); G, ear of locust from within (after Graber). All highly magnified.

Many of the numerous hairs (fig. 9, E) that cover the body of an insect have a tactile function. The sense of smell resides chiefly in the feelers, on whose segments occur tiny pits, often guarded by peg-like or tooth-like structures and containing rod-like cells (fig. 9, C) in connexion with large nerve-cells. It is said that 13,000 such olfactory organs are present on the feeler of a wasp, and 40,000 on the complex antennae of a male cockchafer. Organs of similar type on the maxillae and epipharynx appear to exercise the function of taste.

After Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 10.—Dorsal Muscles, Heart and Pericardial Tendons of Cockroach.

Muscular System.—The muscles in the Hexapoda are striated, as in Arthropods generally, the large fibres being associated in bundles which are attached from point to point of the cuticle, so as to move adjacent sclerites with respect to one another (see figs. 8, 10). For example, the contraction of the tergo-sternal muscles, connecting the dorsal with the ventral sclerites of the abdomen, lessens the capacity of the abdominal region, while the contraction of the powerful muscles arising from the thoracic walls, and inserted into the proximal ends of the thighs, flexes or extends the legs.

Circulatory System.—Insects afford an excellent illustration of the remarkable type of blood-system characterizing the Arthropoda. The dorsal vessel is an elongate tube, whose abdominal portion is usually chambered, forming a contractile heart (fig. 10). At the constrictions between the chambers are paired slits, through which the blood passes from the surrounding pericardial sinus. The dorsal vessel is prolonged anteriorly into an aorta, through which the blood is propelled into the great body-cavity or haemocoel. After bathing the various tissues and organs, the blood returns dorsalwards into the pericardial sinus through fine perforations of its floor, and so makes its way into the heart again. Some water-bugs, e.g. of the families Belostomatidae, Nepidae, Corixidae and Hydrometridae have a pulsating sac at each knee-joint to assist the flow of blood through the legs, while in dragon-flies and locusts (Acridiidae) there is a ventral pulsating diaphragm, which forms the roof of a sinus enclosing the nerve-cords.

After Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 11.—Ventral Portion of Air-Tubes in Cockroach.

Respiratory System.—As mentioned above, respiration by means of air-tubes (tracheae) is a most characteristic feature of the Hexapoda. An air-tube consists of an epithelium of large polygonal cells with a thin basement-membrane externally and a chitinous layer internally, the last-named being continuous with the outer cuticle. The chitinous layer is usually strengthened by thread-like thickenings which, in the region close to the outer opening of the tube, form a network enclosing polygonal areas, but which, through most of the tracheal system, are arranged spirally, the strengthening thread not forming a continuous spiral, but being interrupted after a few turns around the tube. The tracheal system in Hexapods is very complex, forming a series of longitudinal trunks with transverse anastomosing connexions (fig. 11), and extending by the finest sub-division and by repeated branching into all parts of the body. In insects of active flight the tubes swell out into numerous air-sacs, by which the breathing capacity is much increased.

Atmospheric air gains access to the air-tubes through paired spiracles or stigmata, which usually occur laterally on most of the body-segments. These spiracles have firm chitinous edges, and can be closed by valves moved by special muscles. When the spiracles are open and the body contracts, air is expired. The subsequent expansion of the body causes fresh air to enter the tracheal system, and if the spiracles be then closed and the body again contracted, this air is driven to the finest branches of the air-tubes, where a direct oxygenation of the tissues takes place. The physiology of respiration has been carefully studied by F. Plateau (1884). In aquatic insects various devices for obtaining or entangling air are found; these modifications are described in the special articles on the various orders of insects (Coleoptera, Hemiptera, &c.). Many insects have aquatic larvae, some of which take in atmospheric air at intervals, while others breathe dissolved air by means of tracheal gills. These modifications are mentioned below in the section on metamorphosis.

From Miall and Denny, The Cockroach, Lovell Reeve & Co.
Fig. 12.—Food Canal of Cockroach.
s,	Salivary glands and reservoir.
c,	Crop (the gizzard below it).
coe,	Caecal tubes (below them the stomach).
k,	Kidney tubes.
i,	Intestine.
r,	Rectum.

Digestive System.—A striking feature in the food-canal of the Hexapoda, as in other Arthropods, is the great extent of the “fore-gut” and “hind-gut,” lined with a chitinous cuticle, continuous with the exoskeleton. The fore-gut is composed of a tubular gullet, a large sac-like crop (fig. 12, c) and a proventriculus or “gizzard,” whose function is to strain the food-substances before they pass on into the tubular stomach, which has no chitinous lining. This organ, usually regarded as a “mid-gut,” gives off a number of secretory caecal tubes (fig. 12, coe). At its hinder end it is continuous with the hind-gut, which is usually differentiated into a tubular coiled intestine (fig. 12, i) and a swollen rectum (fig. 12, r). From the fore-end of the hind-gut arise the slender Malpighian tubes (fig. 12, k), which have a renal function.

On either side of the gullet are from one to ten pairs of salivary glands (fig. 12, s) whose ducts open into the mouth. Some of these glands may be modified for special purposes—as silk-producing glands in caterpillars or as poison-glands in blood-sucking flies and bugs. The food passing into the crop is there acted on by the saliva and also by an acid gastric juice which passes forwards from the stomach through the proventriculus. As the various portions of the food undergo digestion, they are allowed to pass through the proventriculus into the stomach, where the nutrient substances are absorbed.

Excretory System.—Nitrogenous waste-matter is removed from the body by the Malpighian tubes which open into the food-canal, usually where the hind-gut joins the stomach. These tubes vary in number from four to over a hundred in different orders of insects. The cells which line them and also the cavities of the tubes contain urates, which are excreted from the blood in the surrounding body-cavity. This cavity contains an irregular mass of whitish tissue, the fat-body, consisting of fat-cells which undergo degradation and become more or less filled with urates. When the worn-out cells are broken down, the urates are carried dissolved in the blood to the Malpighian tubes for excretion. The fat-body is therefore the seat of important metabolic processes in the hexapod body.

Reproductive System.—All the Hexapoda are of separate sexes. The ovaries (fig. 13) in the female are paired, each ovary consisting of a variable number of tubes (one in the bristle-tail Campodea and fifteen hundred in a queen termite) in which the eggs are developed. From each ovary an oviduct (fig. 13, od) leads, and in some of the more primitive insects (bristle-tails, earwigs, may-flies) the two oviducts open separately direct to the exterior. Usually they open into a median vagina, formed by an ectodermal inpushing and lined with chitin. The vagina usually opens in front of the eighth abdominal sternite. Behind it is situated a spermatheca (fig. 14, sp) and the ovipositor previously mentioned, with its three pairs of processes (Fig. 14, G, g).

From Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 13.—Ovaries of Cockroach, with Oviducts Od and Colleterial Glands CG.

From Miall and Denny, The Cockroach, Lovell Reeve & Co.
Fig. 14.—Hinder Abdominal Segment and Ovipositor of Female Cockroach. Magnified.
T8 &c.	Tergites.
S7,	7th Sternite.
S8,	Sclerite between 7th and 8th sterna.
S9,	8th Sclerite.
Od,	Vagina.
sp,	Spermatheca.
G,	Anterior, and g, posterior gonapophyses.

The paired testes of the male consist of a variable number of seminal tubes, those of each testis opening into a vas deferens. In some bristle-tails and may-flies, the two vasa deferentia open separately, but usually they lead into a sperm-reservoir, whence issues a median ejaculatory duet. The male opening is on the ninth abdominal segment, to which belong the processes that form the claspers or genital armature. Accessory glands are commonly present in connexion both with the male and the female reproductive organs. The poison-glands of the sting in wasps and bees are well-known examples of these.

Embryology

The Egg.—Among the Hexapoda, as in Arthropods generally, the egg is large, containing an accumulation of yolk for the nourishment of the growing embryo. Most insect eggs are of an elongate oval shape; some are globular, others flattened, while others again are flask-shaped, and the outer envelope (chorion) is often beautifully sculptured (figs. 20, d; 21, a, b). Various devices are adopted for the protection of the eggs from mechanical injury or from the attacks of enemies, and for fixing them in appropriate situations. For example, the egg may be raised above the surface on which it is laid by an elongate stalk; the eggs may be protected by a secretion, which in some cases forms a hard protective capsule or “purse”; or they may be covered with shed hairs of the mother, while among water-insects a gelatinous envelope, often of rope-like form, is common. In various groups of the Hexapoda—aphids and some flesh-flies (Sarcophaga), for example—the egg undergoes development within the body of the mother, and the young insect is born in an active state; such insects are said to be “viviparous.”

Parthenogenesis.—A number of cases are known among the Hexapoda of the development of young from the eggs of virgin females. In insects so widely separated as bristle-tails and moths this occurs occasionally. In certain gall-flies (Cynipidae) no males are known to exist at all, and the species seems to be preserved entirely by successive parthenogenetic generations. In other gall-flies and in aphids we find that a sexual generation alternates with one or with many virgin generations. The offspring of the virgin females are in most of these instances females; but among the bees and wasps parthenogenesis occurs normally and always results in the development of males, the “queen” insect laying either a fertilized or unfertilized egg at will.

Maturation, Fertilization and Segmentation.—Polar bodies were first observed in the eggs of Hexapoda by F. Blochmann in 1887. The two nuclei are successively divided from the egg nucleus in the usual way, but they frequently become absorbed in the peripheral protoplasm instead of being extruded from the egg-cell altogether. It appears that in parthenogenetic eggs two polar nuclei are formed. According to A. Petrunkevich (1901–1903), the second polar nucleus uniting with one daughter-nucleus of the first polar body gives rise to the germ-cells of the parthenogenetically-produced male. There is no reunion of the second polar nucleus with the female pronucleus, but, according to the recent work of L. Doncaster (1906–1907) on the eggs of sawflies, the number of chromosomes is not reduced in parthenogenetic egg-nuclei, while, in eggs capable of fertilization, the usual reduction-divisions occur. Fertilization takes place as the egg is laid, the spermatozoa being ejected from the spermatheca of the female and making their way to the protoplasm of the egg through openings (micropyles) in its firm envelope. The segmentation of the fertilized nucleus results in the formation of a number of nuclei which arrange themselves around the periphery of the egg and, the protoplasm surrounding them becoming constricted, a blastoderm or layer of cells, enclosing the central yolk, is formed. Within the yolk the nuclei of some “yolk cells” can be distinguished.

From Nussbaum in Miall and Denny’s, The Cockroach, Lovell, Reeve & Co.

Fig. 15.—Diagram showing Formation of Germinal Layers. E, ectoderm; M, inner layer. Magnified.

Germinal Layers and Food-Canal.—The embryo begins to develop as an elongate, thickened, ventral region of the blastoderm which is known as the ventral plate or germ band. Along this band a median furrow appears, and a mass of cells sinks within, the one-layered germ band thus becoming transformed into a band of two cell-layers (fig. 15). In some cases the inner layer is formed not by invagination but by proliferation or by delamination. The outer of these two layers (fig. 15, E) is the ectoderm. With regard to the inner layer (endoblast of some authors, fig. 15, M) much difference of opinion has prevailed. It has usually been regarded as representing both endoderm and mesoderm, and the groove which usually leads to its formation has been compared to the abnormally elongated blastopore of a typical gastrula. No doubt can be entertained that the greater part of the inner layer corresponds to the mesoderm of more ordinary embryos, for the coelomic pouches, the germ-cells, the musculature and the vascular system all arise from it. Further, there is general agreement that the chitin-lined fore-gut and hind-gut, which form the greater part of the digestive tract, arise from ectodermal invaginations (stomodaeum and proctodaeum respectively) at the positions of the future mouth and anus. The origin of the mid-gut (mesenteron), that has no chitinous lining in the developed insect, is the disputed point. According to the classical researches of A. Kowalevsky (1871 and 1887) on the embryology of the water-beetle Hydrophilus and of the muscid flies, an anterior and a posterior endoderm-rudiment both derived from the “endoblast” become apparent at an early stage, in close association with the stomodaeum and the proctodaeum respectively. These two endoderm-rudiments ultimately grow together and give rise to the epithelium of the mid-gut. These results were confirmed by the observations of K. Heider and W. M. Wheeler (1889) on the embryos of two beetles—Hydrophilus and Doryphora respectively. V. Graber, however (1889), stated that in the Muscidae, while the anterior endoderm-rudiment arises as Kowalevsky had observed, the posterior part of the “mid-gut” has its origin as a direct outgrowth from the proctodaeum. The recent researches of R. Heymons (1895) on the Orthoptera, and of A. Lécaillon (1898) on various leaf beetles, tend to show that the whole of the “mid-gut” arises from the proliferation of cells at the extremity of the stomodaeum and of the proctodaeum. On this view the entire food-canal in most Hexapoda must be regarded as of ectodermal origin, the “endoblast” represents mesoderm only, and the median furrow whence it arises can be no longer compared with the blastopore. According to Heymons, the yolk-cells must be regarded as the true endoderm in the hexapod embryo, for he states (1897) that in the bristle-tail Lepisma and in dragon-flies they give rise to the mid-gut. These views are not, however, supported by other recent observers. J. Carrière’s researches (1897) on the embryology of the mason bee (Chalicodoma) agree entirely with the interpretations of Kowalevsky and Heider, and so on the whole do those of F. Schwangart, who has studied (1904) the embryonic development of Lepidoptera. He finds that the endoderm arises from an anterior and a posterior rudiment derived from the “endoblast,” that many of the cells of these rudiments wander into the yolk, and that the mesenteric epithelium becomes reinforced by cells that migrate from the yolk. K. Escherich (1901), after a new research on the embryology of the muscid Diptera, claims that the fore and hind endodermal rudiments arise from the blastoderm by invagination, and are from their origin distinct from the mesoderm. On the whole it seems likely that the endoderm is represented in part by the yolk, and in part by those anterior and posterior rudiments which usually form the mesenteron, but that in some Hexapoda the whole digestive tract may be ectodermal. It must be admitted that some or the later work on insect embryology has justified the growing scepticism in the universal applicability of the “germ-layer theory.” Heider has suggested, however, that the apparent origin of the mid-gut from the stomodaeum and proctodaeum may be explained by the presence of a “latent endoderm-group” in those invaginations.

From Nussbaum in Miall and Denny, The Cockroach, Lovell Reeve & Co.

Fig. 16.—Cross section of Embryo of German Cockroach (Phyllodromia). S, serosa; A, amnion; E, ectoderm; N, rudiment of nerve-cord; M, mesodermal pouches.

Embryonic Membranes.—A remarkable feature in the embryonic development of most Hexapoda is the formation of a protective membrane analogous to the amnion of higher Vertebrates and known by the same term. Usually there arises around the edge of the germ band a double fold in the undifferentiated blastoderm, which grows over the surface of the embryo, so that its inner and outer layers become continuous, forming respectively the amnion and the serosa (fig. 16, A, S). The embryo of a moth, a dragon-fly or a bug is invaginated into the yolk at the head end, the portion of the blastoderm necessarily pushed in with it forming the amnion. The embryo thus becomes transferred to the dorsal face of the egg, but at a later stage it undergoes reversion to its original ventral position. In some parasitic Hymenoptera there is only a single embryonic membrane formed by delamination from the blastoderm, while in a few insects, including the wingless spring-tails, the embryonic membranes are vestigial or entirely wanting. In the bristle-tails Lepisma and Machilis, an interesting transitional condition of the embryonic membranes has lately been shown by Heymons. The embryo is invaginated into the yolk, but the surface edges of the blastoderm do not close over, so that a groove or pore puts the insunken space that represents the amniotic cavity into communication with the outside. Heymons believes that the “dorsal organ” in the embryos of the lower Arthropoda corresponds with the region invaginated to form the serosa of the hexapod embryo. Wheeler, however, compares with the “dorsal organ” the peculiar extra embryonic membrane or indusium which he has observed between serosa and amnion in the embryo of the grasshopper Xiphidium.

Metameric Segmentation.—The segments are perceptible at a very early stage of the development as a number of transverse bands arranged in a linear sequence. The first segmentation of the ventral plate is not, however, very definite, and the segmentation does not make its appearance simultaneously throughout the whole length of the plate; the anterior parts are segmented before the posterior. In Orthoptera and Thysanura, as well as some others of the lower insects, twenty-one of these divisions—not, however, all similar—may be readily distinguished, six of which subsequently enter into the formation of the head, three going to the thorax and twelve to the abdomen. In Hemiptera only eleven and in Collembola only six abdominal segments have been detected. The first and last of these twenty-one divisions are so different from the others that they can scarcely be considered true segments.

Head Segments.—In the adult insect the head is insignificant in size compared with the thorax or abdomen, but in the embryo it forms a much larger portion of the body than it does in the adult. Its composition has been the subject of prolonged difference of opinion. Formerly it was said that the head consisted of four divisions, viz. three segments and the procephalic or prae-oral lobes. It is now ascertained that the procephalic lobes consist of three divisions, so that the head must certainly be formed from at least six segments. The first of these, according to the nomenclature of Heymons (see fig. 17), is the mouth or oral piece; the second, the antennal segment; the third, the intercalary or prae-mandibular segment; while the fourth, fifth, and sixth are respectively the segments of the mandibles and of the first and second maxillae. These six divisions of the head are diverse in kind, and subsequently undergo so much change that the part each of them takes in the formation of the head-capsule is not finally determined. The labrum and clypeus are developed as a single prolongation of the oral piece, not as a pair of appendages. The antennal segment apparently entirely disappears, with the exception of a pair of appendages it bears; these become the antennae; it is possible that the original segment, or some part of it, may even become a portion of the actual antennae. The intercalary segment has no appendages, nor rudiments thereof, except, according to H. Uzel (1897), in the thysanuran Campodea, and probably entirely disappears, though J. H. Comstock and C. Kochi believe that the labrum belongs to it. The appendages of the posterior three or trophal segments become the parts of the mouth. The appendages of the two maxillary segments arise as treble instead of single projections, thus differing from other appendages. From these facts it appears that the anterior three divisions of the head differ strongly from the posterior three, which greatly resemble thoracic segments; hence it has been thought possible that the anterior divisions may represent a primitive head, to which three segments and their leg-like appendages were subsequently added to form the head as it now exists. This is, however, very doubtful, and an entirely different inference is possible. Besides the five limb-bearing somites just enumerated, two others must now be recognized in the head. One of these is the ocular segment, in front of the antennal, and behind the primitive pre-oral segment. The other is the segment of the maxillulae (see above, under Jaws), behind the mandibular somite; the presence of this in the embryo of the collembolan Anurida has been lately shown (1900) by J. W. Folsom (fig. 18, v. 5), who terms the maxillulae “superlinguae” on account of their close association with the hypopharynx or lingua. In reference to the structure of the head-capsule in the imago, it appears that the clypeus and labrum represent, as already said, an unpaired median outgrowth of the oral piece. According to W. A. Riley (1904) the epicranium or “vertex,” the compound eyes and the front divisions of the genae are formed by the cephalic lobes of the embryo (belonging to the ocular segment), while the mandibular and maxillary segments form the hinder parts of the genae and the hypopharynx.

After Heymons.

Fig. 17.—Morphology of an Insect: the embryo of Gryllotalpa, somewhat diagrammatic. The longitudinal segmented band along the middle line represents the early segmentation of the nervous system and the subsequent median field of each sternite; the lateral transverse unshaded bands are the lateral fields of each segment; the shaded areas indicate the more internally placed mesoderm layer. The segments are numbered 1-21; 1-6 will form the head, 7-9 the thorax, 10-21 the abdomen. A, anus; Abx₁ Abx₁₁, appendage of 1st and of 11th abdominal segments; Ans, anal piece = telson or 12th abdominal segment; Ant, antenna; De, deuterencephalon; Md, mandible; Mx₁, first maxilla; Mx₂, second maxilla or labium; O, mouth; Obcl, rudimentary labrum and clypeus; Pre, protencephalon; St₁ St₁₀, stigmata 1 and 10; Terg, tergite; Thx₁, appendage of first thoracic segment; Tre, tritencephalon; Ul, a thickening at hinder margin of the mouth.

Great difference of opinion exists as to the hypopharynx, which has even been thought to represent a distinct segment, or the pair of appendages of a distinct segment. Heymons considers that it represents the sternites of the three trophal segments, and that the gula is merely a secondary development. Folsom looks on the hypopharynx as a secondary development. Riley holds that the hypopharynx belongs to the mandibular and maxillary segments, while the cervical sclerites or gula represent the sternum of the labial segment. The ganglia of the nervous system offer some important evidence as to the morphology of the head, and are alluded to below.

Thoracic Segments.—These are always three in number. The three pairs of legs appear very early as rudiments. Though the thoracic segments bear the wings, no trace of these appendages exists till the close of the embryonic life, nor even, in many cases, till much later. The thoracic segments, as seen in an early stage of the ventral plate, display in a well-marked manner the essential elements of the insect segment. These elements are a central piece or sternite, and a lateral field on each side bearing the leg-rudiment. The external part of the lateral field subsequently grows up, and by coalescence with its fellow forms the tergite or dorsal part of the segment.

Abdominal Segments and Appendages.—We have already seen that in numerous lower insects the abdomen is formed from twelve divisions placed in linear fashion. Eleven of these may perhaps be considered as true segments, but the twelfth or terminal one is different, and is called by Heymons a telson; in it is placed the anal orifice, and the mass subsequently becomes the upper and lower laminae anales. In Hemiptera this telson is absent, and the anal orifice is placed quite at the termination of the eleventh segment. Moreover, in this order the abdomen shows at first a division into only nine segments and a terminal mass, which last subsequently becomes divided into two. The appendages of the abdomen are called cerci, stylets and gonapophyses. They differ much according to the kind of insect, and in the adult according to sex. Difference of opinion as to the nature of the abdominal appendages prevails. The cerci, when present, appear in the mature insect to be attached to the tenth segment, but according to Heymons they are really appendages of the eleventh segment, their connexion with the tenth being secondary and the result of considerable changes that take place in the terminal segments. It has been disputed whether any true cerci exist in the higher insects, but they are probably represented in the Diptera and in the scorpion-flies (Mecaptera). In those insects in which a median terminal appendage exists between the two cerci this is considered to be a prolongation of the eleventh tergite. The stylets, when present, are placed on the ninth segment, and in some Thysanura exist also on the eighth segment; their development takes place later in life than that of the cerci. The gonapophyses are the projections near the extremity of the body that surround the sexual orifices, and vary extremely according to the kind of insect. They have chiefly been studied in the female, and form the sting and ovipositor, organs peculiar to this sex. They are developed on the ventral surface of the body and are six in number, one pair arising from the eighth ventral plate and two pairs from the ninth. This has been found to be the case in insects so widely different as Orthoptera and Aculeate Hymenoptera. The genital armature of the male is formed to a considerable extent by modifications of the segments themselves. The development of the armature has been little studied, and the question whether there may be present gonapophyses homologous with those of the female is open.

A. After Wheeler, Journ. Morph. vol. viii., and Folsom, Bull. Mus. Harvard, xxxvi.
B. After Folsom.

Fig. 18.—Embryos of Springtail (Anuridamaritima). Magnified. A, Head-region of germ band. B, Section through head and thorax. The neuromeres are shown in Arabic, the appendages in Roman numerals.

 1, Ocular segment.
 2, Antennal.
 3, Trito-cerebral.
 4, Mandibular.
 5, Maxillular.
 6, Maxillary.
 7, Labial.
 8, Prothoracic.
 9, Mesothoracic.
10, Metathoracic.

In the adult state no insect possesses more than six legs, and they are always attached to the thorax; in many Thysanura there are, however, processes on the abdomen that, as to their position, are similar to legs. In the embryos of many insects there are projections from the segments of the abdomen similar, to a considerable extent, to the rudimentary thoracic legs. The question whether these projections can be considered an indication of former polypody in insects has been raised. They do not long persist in the embryo, but disappear, and the area each one occupied becomes part of the sternite. In some embryos there is but a single pair of these rudiments (or vestiges) situate on the first abdominal segment, and in some cases they become invaginations of a glandular nature. Whether cerci, stylets and gonapophyses are developed from these rudiments has been much debated. It appears that it is possible to accept cerci and stylets as modifications of the temporary pseudopods, but it is more difficult to believe that this is the case with the gonapophyses, for they apparently commence their development considerably later than cerci and stylets and only after the apparently complete disappearance of the embryonic pseudopods. The fact that there are two pairs of gonapophyses on the ninth abdominal segment would be fatal to the view that they are in any way homologous with legs, were it not that there is some evidence that the division into two pairs is secondary and incomplete. But another and apparently insuperable objection may be raised—that the appendages of the ninth segment are the stylets, and that the gonapophyses cannot therefore be appendicular. The pseudopods that exist on the abdomen of numerous caterpillars may possibly arise from the embryonic pseudopods, but this also is far from being established.

Nervous System.—The nervous system is ectodermal in origin, and is developed and segmented to a large extent in connexion with the outer part of the body, so that it affords important evidence as to the segmentation thereof. The continuous layer of cells from which the nervous system is developed undergoes a segmentation analogous with that we have described as occurring in the ventral plate; there is thus formed a pair of contiguous ganglia for each segment of the body, but there is no ganglion for the telson. The ganglia become greatly changed in position during the later life, and it is usually said that there are only ten pairs of abdominal ganglia even in the embryo. In Orthoptera, Heymons has demonstrated the existence of eleven pairs, the terminal pair becoming, however, soon united with the tenth. The nervous system of the embryonic head exhibits three ganglionic masses, anterior to the thoracic ganglionic masses; these three masses subsequently amalgamate and form the sub-oesophageal ganglion, which supplies the trophal segments. In front of the three masses that will form the sub-oesophageal ganglion the mass of cells that is to form the nervous system is very large, and projects on each side; this anterior or “brain” mass consists of three lobes (the prot-, deut-, and tritencephalon of Viallanes and others), each of which might be thought to represent a segmental ganglion. But the protocerebrum contains the ganglia of the ocular segment in addition to those of the procephalic lobes. These three divisions subsequently form the supra-oesophageal ganglion or brain proper. There are other ganglia in addition to those of the ventral chain, and Janet supposes that the ganglia of the sympathetic system indicate the existence of three anterior head-segments; the remains of the segments themselves are, in accordance with this view, to be sought in the stomodaeum. Folsom has detected in the embryo of Anurida a pair of ganglia (fig. 18, 5) belonging to the maxillular (or superlingual) segment, thus establishing seven sets of cephalic ganglia, and supporting his view as to the composition of the head.

Air-tubes.—The air-tubes, like the food-canal, are formed by invaginations of the ectoderm, which arise close to the developing appendages, the rudimentary spiracles appearing soon after the budding limbs. The pits leading from these lengthen into tubes, and undergo repeated branching as development proceeds.

Dorsal Closure.—The germ band evidently marks the ventral aspect of the developing insect, whose body must be completed by the extension of the embryo so as to enclose the yolk dorsally. The method of this dorsal closure varies in different insects. In the Colorado beetle (Doryphora), whose development has been studied by W. M. Wheeler, the amnion is ruptured and turned back from covering the germ band, enclosing the yolk dorsally and becoming finally absorbed, as the ectoderm of the germ band itself spreads to form the dorsal wall. In some midges and in caddis-flies the serosa becomes ruptured and absorbed, while the germ band, still clothed with the amnion, grows around the yolk. In moths and certain saw-flies there is no rupture of the membranes; the Russian zoologists Tichomirov and Kovalevsky have described the growth of both amnion and embryonic ectoderm around the yolk, the embryo being thus completely enclosed until hatching time by both amnion and serosa. V. Graber has described a similar method of dorsal closure in the saw-fly Hylotoma.

After Heymons, Zeit. Wiss. Zoolog. vol. 53.
Fig. 19.—Cross sections through Abdomen of German Cockroach Embryo. A (later than fig. 16) magnified. B (still more advanced, dorsal closure complete) magnified.
ec,	Ectoderm.
en,	Endoderm.
sp,	Splanchnic layer of mesoderm.
y,	Yolk.
h,	Heart.
p,	Pericardial septum.
c,	Coelom.
g,	Germ-cells surrounded by rudiment-cells of ovarian tubes.
m,	Muscle-rudiment.
n,	Nerve-chain.
f,	Fat body.
s,	Inpushing of ectoderm to form air-tubes.
x,	Secondary body-cavity.

Mesoderm, Coelom and Blood-System.—From the mesoderm most of the organs of the body—muscular, circulatory, reproductive—take their origin. The mass of cells undergoes segmentation corresponding with the outer segmentation of the embryo, and a pair of cavities—the coelomic pouches (fig. 16, M)—are formed in each segment. Each coelomic pouch—as traced by Heymons in his study on the development of the cockroach (Phyllodromia)—divides into three parts, of which the most dorsal contains the primitive germ-cells, the median disappears, and the ventral loses its boundaries as it becomes filled up with the growing fat body (fig. 19). This latter, as well as the heart and the walls of the blood spaces, arises by the modification of mesodermal cells, and the body cavity is formed by the enlargement and coalescence of the blood channels and by the splitting of the fat body. It is therefore a haemocoel, the coelom of the developed insect being represented only by the cavities of the genital glands and their ducts.

Reproductive Organs.—In the cockroach embryo, before the segmentation of the germ-band has begun, the primitive germ-cells can be recognized at the hinder end of the mesoderm, from whose ordinary cells they can be distinguished by their larger size. At a later stage further germ-cells arise from the epithelium of the coelomic pouches from the second to the seventh abdominal segments, and become surrounded by other mesoderm cells which form the ovarian or testicular tubes and ducts (fig. 19, g). In the male of Phyllodromia the rudiment of a vestigial ovary becomes separated from the developing testis, indicating perhaps an originally hermaphrodite condition. An exceedingly early differentiation of the primitive germ-cells occurs in certain Diptera. E. Metchnikoff observed (1866) in the development of the parthenogenetic eggs produced by the precocious larva of the gall-midge Cecidomyia that a large “polar-cell” appeared at one extremity during the primitive cell-segmentation. This by successive divisions forms a group of four to eight cells, which subsequently pass through the blastoderm, and dividing into two groups become symmetrically arranged and surrounded by the rudiments of the ovarian tubes. E. G. Balbiani and R. Ritter (1890) have since observed a similar early origin for the germ-cells in the midge Chironomus and in the Aphidae.

The paired oviducts and vasa deferentia are, as we have seen, mesodermal in origin. The median vagina, spermatheca and ejaculatory duct are, on the other hand, formed by ectodermal inpushings. The classical researches of J. A. Palmén (1884) on these ducts have shown that in may-flies and in female earwigs the paired mesodermal ducts open directly to the exterior, while in male earwigs there is a single mesodermal duct, due either to the coalescence of the two or to the suppression of one. In the absence of the external ectodermal ducts usual in winged insects, these two groups resemble therefore the primitive Aptera. The presence of rudiments of the genital ducts of both sexes in the embryo of either sex is interesting and suggestive. The ejaculatory duct which opens on the ninth abdominal sternum in the adult male arises in the tenth abdominal embryonic segment and subsequently moves forward.

Growth and Metamorphosis

After Marlatt, Ent. Bull. 4, n. s. (U.S. Dept. Agr.).

Fig. 20.—a, Bed-bug (Cimex lectularis, Linn.); newly hatched young from beneath; b, from above; d, egg, magnified; c, foot with claws; e, serrate spine, more highly magnified.

From Mally, Ent. Bull. 24 (U.S. Dept. Agr.).

Fig. 21.—e, f, Owl moth (Heliothis armigera); a, b, egg, highly magnified; c, larva or caterpillar; d, pupa in earthen cell.

After hatching or birth an insect undergoes a process of growth and change until the adult condition is reached. The varied details of this post-embryonic development furnish some of the most interesting facts and problems to the students of the Hexapoda. Wingless insects, such as spring-tails and lice, make their appearance in the form of miniature adults. Some winged insects—cockroaches, bugs (fig. 20) and earwigs, for example—when young closely resemble their parents, except for the absence of wings. On the other hand, we find in the vast majority of the Hexapoda a very marked difference between the perfect insect (imago) and the young animal when newly hatched and for some time after hatching. From the moth’s egg comes a crawling caterpillar (fig. 21, c), from the fly’s a legless maggot (fig. 25, a). Such a young insect is a larva—a term used by zoologists for young animals generally that are decidedly unlike their parents. It is obvious that the hatching of the young as a larva necessitates a more or less profound transformation or metamorphosis before the perfect state is attained. Usually this transformation comes with apparent suddenness, at the penultimate stage of the insect’s life-history, when the passive pupa (fig. 21, d) is revealed, exhibiting the wings and other imaginal structures, which have been developed unseen beneath the cuticle of the larva. Hexapoda with this resting pupal stage in their life-history are said to undergo “a complete transformation,” to be metabolic, or holometabolic, whereas those insects in which the young form resembles the parent are said to be ametabolic. Such insects as dragon-flies and may-flies, whose young, though unlike the parent, develop into the adult form without a resting pupal stage are said to undergo an “incomplete transformation” or to be hemimetabolic. The absence of the pupal stage depends upon the fact that in the ametabolic and hemimetabolic Hexapoda the wing-rudiments appear as lateral outgrowths (fig. 22) of the two hinder thoracic segments and are visible externally throughout the life-history, becoming larger after each moult or casting of the cuticle. Hence, as has been pointed out by D. Sharp (1898), the marked divergence among the Hexapoda, as regards life-history, is between insects whose wings develop outside the cuticle (Exopterygota) and those whose wings develop inside the cuticle (Endopterygota), becoming visible only when the casting of the last larval cuticle reveals the pupa. Metamorphosis among the Hexapoda depends upon the universal acquisition of wings during post-embryonic development—no insect being hatched with the smallest external rudiments of those organs—and on the necessity for successive castings or “moults” (ecdyses) of the cuticle.

After Howard, Insect Life, vol. vii.

Fig. 22.—Nymph of Locust (Schistocera americana), showing wing-rudiments.

Ecdysis.—The embryonic ectoderm of an insect consists of a layer of cells forming a continuous structure, the orifices in it—mouth, spiracles, anus and terminal portions of the genital ducts—being invaginations of the outer wall. This cellular layer is called the hypodermis; it is protected externally by a cuticle, a layer of matter it itself excretes, or in the excretion of which it plays, at any rate, an important part. The cuticle is a dead substance, and is composed in large part of chitin. The cuticle contrasts strongly in its nature with the hypodermis it protects. It is different in its details in different insects and in different stages of the life of the same insect. The “sclerites” that make up the skeleton of the insect (which skeleton, it should be remembered, is entirely external) are composed of this chitinous excretion. The growth of an insect is usually rapid, and as the cuticle does not share therein, it is from time to time cast off by moulting or ecdysis. Before a moult actually occurs the cuticle becomes separated from its connexion with the underlying hypodermis. Concomitant with this separation there is commencement of the formation of a new cuticle within the old one, so that when the latter is cast off the insect appears with a partly completed new cuticle. The new instar—or temporary form—is often very different from the old one, and this is the essential fact of metamorphosis. Metamorphosis is, from this point of view, the sum of the changes that take place under the cuticle of an insect between the ecdyses, which changes only become externally displayed when the cuticle is cast off. The hypodermis is the immediate agent in effecting the external changes.

Adapted from Koerschelt and Herder, and Lowne.

Fig. 23.—Diagram showing position of imaginal buds in larva of fly. I., II., III., the three thoracic segments of the larva; 1, 2, 3, buds of the legs of the imago; h, bud of head-lobes; f, of feeler; e of eye; b, brain.

The study of the physiology of ecdysis in its simpler forms has unfortunately been somewhat neglected, investigators having directed their attention chiefly to the cases that are most striking, such as the transformation of a maggot into a fly, or of a caterpillar into a butterfly. The changes have been found to be made up of two sets of processes: histolysis, by which the whole or part of a structure disappears: and histogenesis, or the formation of the new structure. By histolysis certain parts of the hypodermis are destroyed, while other portions of it develop into the new structures. The hypodermis is composed of parts of two different kinds, viz. (1) the larger part of the hypodermis that exists in the maggot or caterpillar and is dissolved at the metamorphosis; (2) parts that remain comparatively quiescent previously, and that grow and develop when the other parts degenerate. These centres of renovation are called imaginal disks or folds. The adult caterpillar may be described as a creature the hypodermis of which is studded with buds that expand and form the butterfly, while the parts around them degenerate. In some insects (e.g. the maggots of the blowfly, Calliphora vomitoria) the imaginal disks are to all appearance completely separated from the hypodermis, with which they are, however, really organically connected by strings or pedicels. This connexion was not at first recognized and the true nature of imaginal disks was not at first perceived, even by Weismann, to whom their discovery in Diptera is due. In other insects the imaginal disks are less completely disconnected from the superficies of the larval hypodermis, and may indeed be merely patches thereof. The number of imaginal disks in an individual is large, upwards of sixty having been discovered to take part in the formation of the outer body of a fly. With regard to the internal organs, we need only say that transformation occurs in an essentially similar manner, by means of a development from centres distributed in the various organs. The imaginal disks for the outer wall of the body, some of them, at any rate, include mesodermal rudiments (from which the muscles are developed) as well as hypodermis. The imaginal disks make their appearance (that is, have been first detected) at very different epochs in the life; their absolute origin has been but little investigated. Pratt has traced them in the sheep-tick (Melophagus) to an early stage of the embryonic life.

Histolysis and Histogenesis.—The process of destruction of the larval tissues was first studied in the forms where metamorphosis is greatest and most abrupt, viz. in the Muscid Diptera. It was found that the tissues were attacked by phagocytic cells that became enlarged and carried away fragments of the tissue; the cells were subsequently identified as leucocytes or blood-cells. Hence the opinion arose that histolysis is a process of phagocytosis. It has, however, since been found that in other kinds of insects the tissues degenerate and break down without the intervention of phagocytes. It has, moreover, been noticed that even in cases where phagocytosis exists a greater or less extent of degeneration of the tissue may be observed before phagocytosis occurs. This process can therefore only be looked on as a secondary one that hastens and perfects the destruction necessary to permit of the accompanying histogenesis. This view is confirmed by the fate of the phagocytic cells. These do not take a direct part in the formation of the new tissue, but it is believed merely yield their surplus acquisitions, becoming ordinary blood-cells or disappearing altogether. As to the nature of histogenesis, nothing more can be said than that it appears to be a phenomenon similar to embryonic growth, though limited to certain spots. Hence we are inclined to look on the imaginal disks as cellular areas that possess in a latent condition the powers of growth and development that exist in the embryo, powers that only become evident in certain special conditions of the organism. What the more essential of these conditions may be is a question on which very little light has been thrown, though it has been widely discussed.

Much consideration has been given to the nature of metamorphosis in insects, to its value to the creatures and to the mode of its origin. Insect metamorphosis may be briefly described as phenomena of development characterized by abrupt changes of appearance and of structure, occurring during the period subsequent to embryonic development and antecedent to the reproductive state. It is, in short, a peculiar mode of growth and adolescence. The differences in appearance between the caterpillar and the butterfly, striking as they are to the eye, do not sufficiently represent the phenomena of metamorphosis to the intelligence. The changes that take place involve a revolution in the being, and may be summarized under three headings: (1) The food-relations of the individual are profoundly changed, an entirely different set of mouth-organs appears and the kind and quantity of the food taken is often radically different. (2) A wingless, sedentary creature is turned into a winged one with superlative powers of aerial movement. (3) An individual in which the reproductive organs and powers are functionally absent becomes one in which these structures and powers are the only reason for existence, for the great majority of insects die after a brief period of reproduction. These changes are in the higher insects so extreme that it is difficult to imagine how they could be increased. In the case of the common drone-fly, Eristalis tenax, the individual, from a sedentary maggot living in filth, without any relations of sex, and with only unimportant organs for the ingestion of its foul nutriment, changes to a creature of extreme alertness, with magnificent powers of flight, living on the products of the flowers it frequents, and endowed with highly complex sexual structures.

After Westwood, Modern Classification.

Fig. 24.—Campodeiform Larva of a Ground-Beetle (Aepus marinus).

Magnified.

Forms of Larva.—The unlikeness of the young insect to its parent is one of the factors that necessitates metamorphosis. It is instructive, further, to trace among metabolic insects an increase in the degree of this dissimilarity. An adult Hexapod is provided with a firm, well-chitinized cuticle and six conspicuous jointed legs. Many larval Hexapods might be defined in similar general terms, unlike as they are to their parents in most points of detail. Examples of such are to be seen in the grubs of may-flies, dragon-flies, lacewing-flies and ground-beetles (fig. 24). This type of active, armoured larva—often bearing conspicuous feelers on the head and long jointed cercopods on the tenth abdominal segment—was styled campodeiform by F. Brauer (1869), on account of its likeness in shape to the bristle-tail Campodea. As an extreme contrast to this campodeiform type, we take the maggot of the house-fly (fig. 25)—a vermiform larva, with soft, white, feebly-chitinized cuticle and without either head-capsule or legs. Between these two extremes, numerous intermediate forms can be traced: the grub (wireworm) of a click-beetle, with narrow elongate well-armoured body, but with the legs very short; the grub of a chafer, with the legs fairly developed, but with the cuticle of all the trunk-segments soft and feebly chitinized; the well-known caterpillar of a moth (fig. 21, e) or saw-fly, with its long cylindrical body, bearing the six shortened thoracic legs and a variable number of pairs of “pro-legs” on the abdomen (this being the eruciform type of larva); the soft, white, wood-boring grub of a longhorn-beetle or of the saw-fly Sirex, with its stumpy vestiges of thoracic legs; the large-headed but entirely legless, fleshy grub of a weevil; and the legless larva, with greatly reduced head, of a bee. The various larvae of the above series, however, have all a distinct head-capsule, which is altogether wanting in the degraded fly maggot. These differences in larval form depend in part on the surroundings among which the larva finds itself after hatching; the active, armoured grub has to seek food for itself and to fight its own battles, while the soft, defenceless maggot is provided with abundant nourishment. But in general we find that elaboration of imaginal structure is associated with degradation in the nature of the larva, eruciform and vermiform larvae being characteristic of the highest orders of the Hexapoda, so that unlikeness between parent and offspring has increased with the evolution of the class.

After Howard, Ent. Bull. 4, n. s. (U.S. Dept. Agr.).

Fig. 25.—Vermiform Larva (maggot) of House-fly (Musca domestica). Magnified. b, spiracle on prothorax; c, protruded head region; d, tail-end with functional spiracles; e, f, head region with mouth hooks protruded; g, hooks retracted; h, eggs. All magnified.

Hypermetamorphosis.—Among a few of the beetles or Coleoptera (q.v.), and also in the neuropterous genus Mantispa, are found life-histories in which the earliest instar is campodeiform and the succeeding larval stages eruciform. These later stages, comprising the greater part of the larval history, are adapted for an inquiline or a parasitic life, where shelter is assured and food abundant, while the short-lived, active condition enables the newly-hatched insect to make its way to the spot favourable for its future development, clinging, for example, in the case of an oil-beetle’s larva, to the hairs of a bee as she flies towards her nest. The presence of the two successive larval forms in the life-history constitutes what is called hypermetamorphosis. Most significant is the precedence of the eruciform by the campodeiform type. In conjunction with the association mentioned above of the most highly developed imaginal with the most degraded larval structure, it indicates clearly that the active, armoured grub preceded the sluggish soft-skinned caterpillar or maggot in the evolution of the Hexapoda.

Nymph.—The term nymph is applied by many writers on the Hexapoda to all young forms of insects that are not sufficiently unlike their parents to be called larvae. Other writers apply the term to a “free” pupa (see infra). It is in wellnigh universal use for those instars of ametabolous and hemimetabolous insects in which the external wing-rudiments have become conspicuous (fig. 27). The mature dragon-fly nymph, for example, makes its way out of the water in which the early stages have been passed and, clinging to some water-plant, undergoes the final ecdysis that the imago may emerge into the air. Like most ametabolic and hemimetabolic Hexapoda, such nymphs continue to move and feed throughout their lives. But examples are not wanting of a more or less complete resting habit during the latest nymphal instar. In some cicads the mature nymph ceases to feed and remains quiescent within a pillar-shaped earthen chamber. The nymph of a thrips-insect (Thysanoptera) is sluggish, its legs and wings being sheathed by a delicate membrane, while the nymph of the male scale-insect rests enclosed beneath a waxy covering.

Sub-imago.—Among the Hexapoda generally there is no subsequent ecdysis nor any further growth after the assumption of the winged state. The may-flies, however, offer a remarkable exception to this rule. After a prolonged aquatic larval and nymphal life-history, the winged insect appears as a sub-imago, whence, after the casting of a delicate cuticle, the true imago emerges.

Pupa.—In the metabolic Hexapoda the resting pupal instar shows externally the wings and other characteristic imaginal organs which have been gradually elaborated beneath the larval cuticle. It is usual to distinguish between the free pupae (fig. 26, b)—of Coleoptera and Hymenoptera, for example—in which the wings, legs and other appendages are not fixed to the trunk, and the obtect pupae (fig. 21, d)—such as may be noticed in the majority of the Lepidoptera—whose appendages are closely and immovably pressed to the body by a general hardening and fusion of the cuticle. In the degree of mobility there is great diversity among pupae. A gnat pupa swims through the water by powerful strokes of its abdomen, while the caddis-fly pupa, in preparation for its final ecdysis, bites its way out of its subaqueous protective case and rises through the water, so that the fly may emerge into the air. Some pupae are thus more active than some nymphs; the essential character of a pupa is not therefore its passivity, but that it is the instar in which the wings first become evident externally. The division of the winged Hexapoda into Exopteryga and Endopteryga is thus again justified.

If we admit that the larva has, in the phylogeny of insects, gradually diverged from the imago, and if we recollect that in the ontogeny the larva has always to become the imago (and of course still does so) notwithstanding the increased difficulty of the transformation, we cannot but recognize that a period of helplessness in which the transformation may take place is to be expected. It is generally considered that this is sufficient as an explanation of the existence of the pupa. This, however, is not the case, because the greater part of the transformation precedes the disclosure of the pupa, which, as L. C. Miall remarks, is structurally little other “than the fly enclosed in a temporary skin.” Moreover, in many insects with imperfect metamorphosis the change from larva or (as the later stage of the larva is called in these cases) nymph to imago is about as great as the corresponding change in the Holometabola, as the student will recognize if he recalls the histories of Ephemeridae, Odonata and male Coccidae. But in none of these latter cases have the wings to be changed from a position inside the body to become external and actively functional organs. The difference between the nymph or false pupa and the true pupa is that in the latter a whole stage is devoted to the perfecting of the wings and body-wall after the wings have become external organs; the stage is one in which no food is or can be taken, however prolonged may be its existence. Amongst insects with imperfect metamorphosis the nearest approximations to the true pupa of the Holometabola are to be found in the sub-imago of Ephemeridae and in the quiescent or resting stages of Thysanoptera, Aleurodidae and Coccidae. A much more thorough appreciation than we yet possess of the phenomena in these cases is necessary in order completely to demonstrate the special characteristics of the holometabolous transformation. But even at present we can correctly state that the true pupa is invariably connected with the transference of the wings from the interior to the exterior of the body. It cannot but suggest itself that this transference was induced by some peculiarity as to formation of cuticle, causing the growth of the wings to be directed inwards instead of outwards. We may remark that fleas possess no wings, but are understood to possess a true pupa. This is a most remarkable case, but unfortunately very little information exists as to the details of metamorphosis in this group.

From Chittenden, Bull. 4 (n.s.) Div. Ent. U.S. Dept. Agr.

Fig. 26.—a, Saw-toothed Grain-Beetle (Silvanus surinamensis); b, pupa; c, larva, magnified—; d, feeler of larva.

Life-Relations.—Only a brief reference can be made here to the fascinating subject of the life-relations of the larva, nymph and pupa, as compared with those of the imago. For details, the reader may consult the special articles on the various orders and groups of insects. A common result of metamorphosis is that the larva and imago differ markedly in their habitat and mode of feeding. The larva may be aquatic, or subterranean, or a burrower in wood, while the imago is aerial. It may bite and devour solid food, while the imago sucks liquids. It may eat roots or refuse, while the imago lives on leaves and flowers. The aquatic habit of many larvae is associated with endless beautiful adaptations for respiration. The series of paired spiracles on most of the trunk-segments is well displayed, as a rule, in terrestrial larvae—caterpillars and the grubs of most beetles, for example. In many aquatic larvae we find that all the spiracles are closed up, or become functionless, except a pair at the hinder end which are associated with some arrangement—such as the valvular flaps of the gnat larva or the telescopic “tail” of the drone-fly larva—for piercing the surface film and drawing periodical supplies of atmospheric air. A similar restriction of the functional spiracles to the tail-end (fig. 25, d) is seen in many larvae of flies (Diptera) that live and feed buried in carrion or excrement. Other aquatic larvae have the tracheal system entirely closed, and are able to breathe dissolved air by means of tubular or leaf-like gills. Such are the grubs of stone-flies, may-flies (fig. 27) and some dragon-flies and midges. An interesting feature is the difference often to be observed between an aquatic larva and pupa of the same insect in the matter of breathing. The gnat larva, for example, breathes at the tail-end, hanging head-downwards from the surface-film. But the pupa hangs from the surface by means of paired respiratory trumpets on the prothorax, the dorsal thoracic surface, where the cuticle splits to allow the emergence of the fly, being thus directed towards the upper air.

From Miall and Denny (after Vayssière), The Cockroach, Lovell Reeve & Co.

Fig. 27.—Nymph of May-fly (Chloeon dipterum), with wing rudiments (a) and tracheal gill-plates (b, b). Magnified—. (The feelers and legs are cut short.)

A marked disproportion between the life-term of larva and imago is common; the former often lives for months or years, while the latter only survives for weeks or days or hours. Generally the larval is the feeding, the imaginal the breeding, stage of the life-cycle. The extreme of this “division of labour” is seen in those insects whose jaws are vestigial in the winged state, when, the need for feeding all behind them, they have but to pair, to lay eggs and to die. The acquisition of wings is the sign of developed reproductive power.

Paedogenesis.—Nevertheless, the function of reproduction is occasionally exercised by larvae. In 1865 N. Wagner made his classical observations on the production of larvae from unfertilized eggs developed in the precociously-formed ovaries of a larval gall-midge (Cecidomyid), and subsequent observers have confirmed his results by studies on insects of the same family and of the related Chironomidae. The larvae produced by this remarkable method (paedogenesis) of virgin-reproduction are hatched within the parent larva, and in some cases escape by the rupture of its body.

Polyembryony.—Occasionally the power of reproduction is thrown still farther back in the life-history, and it is found that from a single egg a large number of embryos may be formed. P. Marchal has (1904) described this power in two small parasitic Hymenoptera—a Chalcid (Encyrtus) which lays eggs in the developing eggs of the small moth Hyponomeuta, and a Proctotrypid (Polygnotus) which infests a gall-midge (Cecidomyid) larva. In the egg of these insects a small number of nuclei are formed by the division of the nucleus, and each of these nuclei originates by division the cell-layers of a separate embryo. Thus a mass or chain of embryos is produced, lying in a common cyst, and developing as their larval host develops. In this way over a hundred embryos may result from a single egg. Marchal points out the analogy of this phenomenon to the artificial polyembryony that has been induced in Echinoderm and other eggs by separating the blastomeres, and suggests that the abundant food-supply afforded by the host-larva is favourable for this multiplication of embryos, which may be, in the first instance, incited by the abnormal osmotic pressure on the egg.

Duration of Life.—The flour-moth (Ephestia kuhniella) sometimes passes through five or six generations in a single year. Although one of the characteristics of insects is the brevity of their adult lives, a considerable number of exceptions to the general rule have been discovered. These exceptions may be briefly summarized as follows: (1) Certain larvae, provided with food that may be adequate in quantity but deficient in nutriment, may live and go on feeding for many years; (2) certain stages of the life that are naturally “resting stages” may be in exceptional cases prolonged, and that to a very great extent; in this case no food is taken, and the activity of the individual is almost nil; (3) the life of certain insects in the adult state may be much prolonged if celibacy be maintained; a female of Cybister roeselii (a large water-beetle) has lived five and a half years in the adult state in captivity. In addition to these abnormal cases, the life of certain insects is naturally more prolonged than usual. The females of some social insects have been known to live for many years. In Tibicen septemdecim the life of the larva extends over from thirteen to seventeen years. The eggs of locusts may remain for years in the ground before hatching; and there may thus arise the peculiar phenomenon of some species of insect appearing in vast numbers in a locality where it has not been seen for several years.

Classification

Number of Species.—It is now considered that 2,000,000 is a moderate estimate of the species of insects actually existing. Some authorities consider this total to be too small, and extend the number to 10,000,000. Upwards of 300,000 species have been collected and described, and at present the number of named forms increases at the rate of about 8000 species per annum. The greater part by far of the insects existing in the world is still quite unknown to science. Many of the species are in process of extinction, owing to the extensive changes that are taking place in the natural conditions of the world by the extension of human population and of cultivation, and by the destruction of forests; hence it is probable that a considerable proportion of the species at present existing will disappear from the face of the earth before we have discovered or preserved any specimens of them. Nevertheless, the constant increase of our knowledge of insect forms renders classification increasingly difficult, for gaps in the series become filled, and while the number of genera and families increases, the distinctions between these groups become dependent on characters that must seem trivial to the naturalist who is not a specialist.

Orders of Hexapoda.—In the present article it is only possible to treat of the division of the Hexapoda into orders and sub-orders and of the relations of these orders to each other. For further classificatory details, reference must be made to the special articles on the various orders. As regards the vast majority of insects, the orders proposed by Linnaeus are acknowledged by modern zoologists. His classification was founded mainly on the nature of the wings, and five of his orders—the Hymenoptera (bees, ants, wasps, &c.), Coleoptera (beetles), Diptera (two-winged flies), Lepidoptera (moths and butterflies), and Hemiptera (bugs, cicads, &c.)—are recognized to-day with nearly the same limits as he laid down. His order of wingless insects (Aptera) included Crustacea, spiders, centipedes and other creatures that now form classes of the Arthropoda distinct from the Hexapoda; it also included Hexapoda of parasitic and evidently degraded structure, that are now regarded as allied more or less closely to various winged insects. Consequently the modern order Aptera comprises only a very small proportion of Linnaeus’s “Aptera”—the spring-tails and bristle-tails, wingless Hexapoda that stand evidently at a lower grade of development than the bulk of the class. The earwigs, cockroaches and locusts, which Linnaeus included among the Coleoptera, were early grouped into a distinct order, the Orthoptera. The great advance in modern zoology as regards the classification of the Hexapoda lies in the treatment of a heterogeneous assembly which formed Linnaeus’s order Neuroptera. The characters of the wings are doubtless important as indications of relationship, but the nature of the jaws and the course of the life-history must be considered of greater value. Linnaeus’s Neuroptera exhibit great diversity in these respects, and the insects included in it are now therefore distributed into a number of distinct orders. The many different arrangements that have been proposed can hardly be referred to in this article. Of special importance in the history of systematic entomology was the scheme of F. Brauer (1885), who separated the spring tails and bristle-tails as a sub-class Apterygogenea from all the other Hexapoda, these forming the sub-class Pterygogenea distributed into sixteen orders. Brauer in his arrangement of these orders laid special stress on the nature of the metamorphosis, and was the first to draw attention to the number of Malpighian tubes as of importance in classification. Subsequent writers have, for the most part, increased the number of recognized orders; and during the last few years several schemes of classification have been published, in the most revolutionary of which—that of A. Handlirsch (1903–1904)—the Hexapoda are divided into four classes and thirty-four orders! Such excessive multiplication of the larger taxonomic divisions shows an imperfect sense of proportion, for if the term “class” be allowed its usual zoological value, no student can fail to recognize that the Hexapoda form a single well-defined class, from which few entomologists would wish to exclude even the Apterygogenea. In several recent attempts to group the orders into sub-classes, stress has been laid upon a few characters in the imago. C. Börner (1904), for example, considers the presence or absence of cerci of great importance, while F. Klapalek (1904) lays stress on a supposed distinction between appendicular and non-appendicular genital processes. A natural system must take into account the nature of the larva and of the metamorphosis in conjunction with the general characters of the imago. Hence the grouping of the orders of winged Hexapoda into the divisions Exopterygota and Endopterygota, as suggested by D. Sharp, is unlikely to be superseded by the result of any researches into minute imaginal structure. Sharp’s proposed association of the parasitic wingless insects in a group Anapterygota cannot, however, be defended as natural; and recent researches into the structure of these forms enables us to associate them confidently with related winged orders. The classification here adopted is based on Sharp’s scheme, with the addition of suggestions from some of the most recent authors—especially Börner and Enderlein.

Class: HEXAPODA.

Sub-class: Apterygota.

Primitively (?) wingless Hexapods with cumacean mandibles, distinct maxillulae, and locomotor abdominal appendages. Without ectodermal genital ducts. Young closely resemble adults.

The sub-class contains a single

Order: Aptera,

which is divided into two sub-orders:

1. Thysanura (Bristle-tails): with ten abdominal segments; number of abdominal appendages variable. Cerci prominent. Developed tracheal system.

2. Collembola (Spring-tails): with six abdominal segments; appendages of the first forming an adherent ventral tube, those of the third a minute “catch,” those of the fourth (fused basally) a “spring.” Tracheal system reduced or absent.

Sub-class: Exopterygota.

Hexapoda mostly with wings, the wingless forms clearly degraded. Maxillulae rarely distinct. No locomotor abdominal appendages. The wing-rudiments develop visibly outside the cuticle. Young like or unlike parents.

Order: Dermaptera.

Biting mandibles; minute but distinct-maxillulae; second maxillae incompletely fused. When wings are present, the fore-wings are small firm elytra, beneath which the delicate hind-wings are complexly folded. Many forms wingless. Genital ducts entirely mesodermal. Cerci always present; usually modified into unjointed forceps. Numerous (30 or more) Malpighian tubes. Young resembling parents.

Includes two families—the Forficulidae or earwigs (q.v.) and the Hemimeridae.

Order: Orthoptera.

Biting mandibles; vestigial maxillulae; second maxillae incompletely fused. Wings usually well developed, net-veined; the fore-wings of firmer texture than the hind-wings, whose anal area folds fanwise beneath them. Jointed cerci always present; ovipositor well developed. Malpighian tubes numerous (100-150). Young resemble parents.

Includes stick and leaf insects, cockroaches, mantids, grasshoppers, locusts and crickets (see Orthoptera).

Order: Plecoptera.

Biting mandibles; second maxillae incompletely fused. Fore-wings similar in texture to hind-wings, whose anal area folds fanwise. Jointed, often elongate, cerci. Numerous (50-60) Malpighian tubes. Young resembling parents, but aquatic in habit, breathing dissolved air by thoracic tracheal gills.

Includes the single family of the Perlidae (Stone-flies), formerly grouped with the Neuroptera.

Order: Isoptera.

Biting mandibles; second maxillae incompletely fused. Fore-wings similar in shape and texture to hind-wings, which do not fold. In most species the majority of individuals are wingless. Short, jointed cerci. Six or eight Malpighian tubes. Young resembling adults; terrestrial throughout life.

Includes two families, formerly reckoned among the Neuroptera—the Embiidae and the Termitidae or “White Ants” (see Termite).

Order: Corrodentia.

Biting mandibles; second maxillae incompletely fused; maxillulae often distinct. Cerci absent. Four Malpighian tubes.

Includes two sub-orders, formerly regarded as Neuroptera:—

1. Copeognatha: Corrodentia with delicate cuticle. Wings usually developed; the fore-wings much larger than the hind-wings. One family, the Psocidae (Book-lice). These minute insects are found amongst old books and furniture.

2. Mallophaga: Parasitic wingless Corrodentia (Bird-lice).

Order: Ephemeroptera.

Jaws vestigial. Fore-wings much larger than hind-wings. Elongate, jointed cerci. Genital ducts paired and entirely mesodermal. Malpighian tubes numerous (40). Aquatic larvae with distinct maxillulae, breathing dissolved air by abdominal tracheal gills. Penultimate instar a flying sub-imago. [Includes the single family of the Ephemeridae or may-flies. See also Neuroptera, in which this order was formerly comprised.]

Order: Odonata.

Biting mandibles. Wings of both pairs closely alike; firm and glassy in texture. Prominent, unjointed cerci, male with genital armature on second abdominal segment. Malpighian tubes numerous (50-60). Aquatic larvae with caudal leaf-gills or with rectal tracheal system.

Includes the three families of dragon-flies. Formerly comprised among the Neuroptera.

Order: Thysanoptera.

Piercing mandibles, retracted within the head-capsule. First maxillae also modified as piercers; maxillae of both pairs with distinct palps. Both pairs of wings similar, narrow and fringed. Four Malpighian tubes. Cerci absent. Ovipositor usually present. Young resembling parents, but penultimate instar passive and enclosed in a filmy pellicle.

Includes three families of Thrips (see Thysanoptera).

Order: Hemiptera.

Mandibles and first maxillae modified as piercers; second maxillae fused to form a jointed, grooved rostrum. Wings usually present. Four Malpighian tubes. Cerci absent. Ovipositor developed.

Includes two sub-orders:—

1. Heteroptera: Rostrum not in contact with haunches of fore-legs. Fore-wings partly coriaceous. Young resembling adults.

Includes the bugs, terrestrial and aquatic.

2. Homoptera: Rostrum in contact with haunches of fore-legs. Fore-wings uniform in texture. Young often larvae. Penultimate instar passive in some cases.

Includes the cicads, aphides and scale-insects (see Hemiptera).

Order: Anoplura.

Piercing jaws modified and reduced, a tubular, protrusible sucking-trunk being developed; mouth with hooks. Wingless, parasitic forms. Cerci absent. Four Malpighian tubes. Young resembling adults.

Includes the family of the Lice (Pediculidae), often reckoned as Hemiptera (q.v.). See also Louse.

Sub-class: Endopterygota.

Hexapoda mostly with wings; the wingless forms clearly degraded or modified. Maxillulae vestigial or absent. No locomotor abdominal appendages (except in certain larvae). Young animals always unlike parents, the wing-rudiments developing beneath the larval cuticle and only appearing in a penultimate pupal instar, which takes no food and is usually passive.

Order: Neuroptera.

Biting mandibles; second maxillae completely fused. Prothorax large and free. Membranous, net-veined wings, those of the two pairs closely alike. Six or eight Malpighian tubes. Cerci absent. Larva campodeiform, usually feeding by suction (exceptionally hypermetamorphic with subsequent eruciform instars). Pupa free.

Includes the alder-flies, ant-lions and lacewing-flies. See Neuroptera.

Order: Coleoptera.

Biting mandibles; second maxillae very intimately fused. Prothorax large and free. Fore-wings modified into firm elytra, beneath which the membranous hind-wings (when present) can be folded. Cerci absent. Four or six Malpighian tubes. Larva campodeiform or eruciform. Pupa free.

Includes the beetles and the parasitic Stylopidae, often regarded as a distinct order (Strepsiptera). (See Coleoptera.)

Order: Mecaptera.

Biting mandibles; first maxillae elongate; second maxillae completely fused. Prothorax small. Two pairs of similar, membranous wings, with predominantly longitudinal neuration. Six Malpighian tubes. Larva eruciform. Pupa free. Cerci present.

Includes the single family of Panorpidae (scorpion-flies), often comprised among the Neuroptera.

Order: Trichoptera.

Mandibles present in pupa, vestigial in imago; maxillae suctorial without specialization; first maxillae with lacinia, galea and palp. Prothorax small. Two pairs of membranous, hair-covered wings, with predominantly longitudinal neuration. Larvae aquatic and eruciform. Pupa free. Six Malpighian tubes. Cerci absent.

Includes the caddis-flies. See Neuroptera, among which these insects were formerly comprised.

Order: Lepidoptera.

Mandibles absent in imago, very exceptionally present in pupa; first maxillae nearly always without laciniae and often without palps, or only with vestigial palps, their galeae elongated and grooved inwardly so as to form a sucking trunk. Prothorax small. Wings with predominantly longitudinal neuration, covered with flattened scales. Fore-wings larger than hind-wings. Cerci absent. Four (rarely 6 or 8) Malpighian tubes. Larvae eruciform, with rarely more than five pairs of abdominal prolegs. Pupa free in the lowest families, in most cases incompletely or completely obtect.

Includes the moths and butterflies. See Lepidoptera.

Order: Diptera.

Mandibles rarely present, adapted for piercing; first maxillae with palps; second maxillae forming with hypopharynx a suctorial proboscis. Prothorax small, intimately united to mesothorax. Fore-wings well developed; hind-wings reduced to stalked knobs (“halteres”). Cerci present but usually reduced. Four Malpighian tubes. Larvae eruciform without thoracic legs, or vermiform without head-capsule. Pupa incompletely obtect or free, and enclosed in the hardened cuticle of the last larval instar (puparium).

Includes the two-winged flies (see Diptera), which may be divided into two sub-orders:—

1. Orthorrhapha: Larva eruciform. Cuticle of pupa or puparium splitting longitudinally down the back, to allow escape of imago.

Comprises the midges, gnats, crane-flies, gad-flies, &c.

2. Cyclorrhapha: Larva vermiform (no head-capsule). Puparium opening by an anterior “lid.”

Comprises the hover-flies, flesh-flies, bot-flies, &c.

Order: Siphonaptera.

Mandibles fused into a piercer; first maxillae developed as piercers; palps of both pairs of maxillae present; hypopharynx wanting. Prothorax large. Wings absent or vestigial. Larva eruciform, limbless.

Includes the fleas.

Order: Hymenoptera.

Biting mandibles; second maxillae incompletely or completely fused; often forming a suctorial proboscis. Prothorax small, and united to mesothorax. First abdominal segment united to metathorax. Wings membranous, fore-wings larger than hind-wings. Ovipositor always well developed, and often modified into a sting. Numerous (20-150) Malpighian tubes (in rare cases, 6-12 only). Larva eruciform, with seven or eight pairs of abdominal prolegs, or entirely legless. Pupa free.

Includes two sub-orders:—

1. Symphyta: Abdomen not basally constricted. Larvae caterpillars with thoracic legs and abdominal prolegs.

Comprises the saw-flies.

2. Apocrita: Abdomen markedly constricted at second segment. Larvae legless grubs.

Comprises gall-flies, ichneumon-flies, ants, wasps, bees. See Hymenoptera.

Geological History

The classification just given has been drawn up with reference to existing insects, but the great majority of the extinct forms that have been discovered can be referred with some confidence to the same orders, and in many cases to recent families. The Hexapoda, being aerial, terrestrial and fresh-water animals, are but occasionally preserved in stratified rocks, and our knowledge of extinct members of the class is therefore fragmentary, while the description, as insects, of various obscure fossils, which are perhaps not even Arthropods, has not tended to the advancement of this branch of zoology. Nevertheless, much progress has been made. Several Silurian fossils have been identified as insects, including a Thysanuran from North America, but upon these considerable doubt has been cast.

The Devonian rocks of Canada (New Brunswick) have yielded several fossils which are undoubtedly wings of Hexapods. These have been described by S. H. Scudder, and include gigantic forms related to the Ephemeroptera.

In the Carboniferous strata (Coal measures) remains of Hexapods become numerous and quite indisputable. Many European forms of this age have been described by C. Brongniart, and American by S. H. Scudder. The latter has established, for all the Palaeozoic insects, an order Palaeodictyoptera, there being a closer similarity between the fore-wings and the hind-wings than is to be seen in most living orders of Hexapoda, while affinities are shown to several of these orders—notably the Orthoptera, Ephemeroptera, Odonata and Hemiptera. It is probable that many of these Carboniferous insects might be referred to the Isoptera, while others would fall into the existing orders to which they are allied, with some modification of our present diagnoses. Of special interest are cockroach-like forms, with two pairs of similar membranous wings and a long ovipositor, and gigantic insects allied to the Odonata, that measured 2 ft. across the outspread wings. A remarkable fossil from the Scottish Coal-measures (Lithomantis) had apparently small wing-like structures on the prothorax, and in allied genera small veined outgrowths—like tracheal gills—occurred on the abdominal segments. To the Permian period belongs a remarkable genus Eugereon, that combines hemipteroid jaws with orthopteroid wing-neuration. With the dawn of the Mesozoic epoch we reach Hexapods that can be unhesitatingly referred to existing orders. From the Trias of Colorado, Scudder has described cockroaches intermediate between their Carboniferous precursors and their present-day descendants, while the existence of endopterygotous Hexapods is shown by the remains of Coleoptera of several families. In the Jurassic rocks are found Ephemeroptera and Odonata, as well as Hemiptera, referable to existing families, some representatives of which had already appeared in the oldest of the Jurassic ages—the Lias. To the Lias also can be traced back the Neuroptera, the Trichoptera, the orthorrhaphous Diptera and, according to the determination of certain obscure fossils, also the Hymenoptera (ants). The Lithographic stone of Kimmeridgian age, at Solenhofen in Bavaria, is especially rich in insect remains, cyclorrhaphous Diptera appearing here for the first time. In Tertiary times the higher Diptera, besides Lepidoptera and Hymenoptera, referable to existing families, become fairly abundant. Numerous fossil insects preserved in the amber of the Baltic Oligocene have been described by G. L. Mayr and others, while Scudder has studied the rich Oligocene faunas of Colorado (Florissant) and Wyoming (Green River). The Oeningen beds of Baden, of Miocene age, have also yielded an extensive insect fauna, described fifty years ago by O. Heer. Further details of the geological history of the Hexapoda will be found in the special articles on the various orders. Fragmentary as the records are, they show that the Exopterygota preceded the Endopterygota in the evolution of the class, and that among the Endopterygota those orders in which the greatest difference exists between imago and larva—the Lepidoptera, Diptera and Hymenoptera—were the latest to take their rise.

Geographical Distribution

The class Hexapoda has a world-wide range, and so have most of its component orders. The Aptera have perhaps the most extensive distribution of all animals, being found in Franz Josef Land and South Victoria Land, on the snows of Alpine glaciers, and in the depths of the most extensive caves. Most of the families and a large proportion of the genera of insects are exceedingly widespread, but a study of the genera and species in any of the more important families shows that faunas can be distinguished whose headquarters agree fairly with the regions that have been proposed to express the distribution of the higher vertebrates. Many insects, however, can readily extend their range, and a careful study of their distribution leads us to discriminate between faunas rather than definitely to map regions. A large and dominant Holoarctic fauna, with numerous subdivisions, ranges over the great northern continents, and is characterized by the abundance of certain families like the Carabidae and Staphylinidae among the Coleoptera and the Tenthredinidae among the Hymenoptera. The southern territory held by this fauna is invaded by genera and species distinctly tropical. Oriental types range far northwards into China and Japan. Ethiopian forms invade the Mediterranean area. Neotropical and distinctively Sonoran insects mingle with members of the Holoarctic fauna across a wide “transition zone” in North America. “Wallace’s line” dividing the Indo-Malayan and Austro-Malayan sub-regions is frequently transgressed in the range of Malayan insects. The Australian fauna is rich in characteristic and peculiar genera, and New Zealand, while possessing some remarkable insects of its own, lacks entirely several families with an almost world-wide range—for example, the Notodontidae, Lasiocampidae, and other families of Lepidoptera. Interesting relationships between the Ethiopian and Oriental, the Neotropical and West African, the Patagonian and New Zealand faunas suggest great changes in the distribution of land and water, and throw doubt on the doctrine of the permanence of continental areas and oceanic basins. Holoarctic types reappear on the Andes and in South Africa, and even in New Zealand. The study of the Hexapoda of oceanic islands is full of interest. After the determination of a number of cosmopolitan insects that may well have been artificially introduced, there remains a large proportion of endemic species—sometimes referable to distinct genera—which suggest a high antiquity for the truly insular faunas.

Relationships and Phylogeny

The Hexapoda form a very clearly defined class of the Arthropoda, and many recent writers have suggested that they must have arisen independently of other Arthropods from annelid worms, and that the Arthropoda must, therefore, be regarded as an “unnatural,” polyphyletic assemblage. The cogent arguments against this view are set forth in the article on Arthropoda. A near relationship between the Apterygota and the Crustacea has been ably advocated by H. J. Hansen (1893). It is admitted on all hands that the Hexapoda are akin to the Chilopoda. Verhoeff has lately (1904) put forward the view that there are really six segments in the hexapodan thorax and twenty in the abdomen—the cerci belonging to the seventeenth abdominal segment thus showing a close agreement with the centipede Scolopendra. On the other hand, G. H. Carpenter (1899, 1902–1904) has lately endeavoured to show an exact numerical correspondence in segmentation between the Hexapoda, the Crustacea, the Arachnida, and the most primitive of the Diplopoda. On either view it may be believed that the Hexapoda arose with the allied classes from a primitive arthropod stock, while the relationships of the class are with the Crustacea, the Chilopoda and the Diplopoda, rather than with the Arachnida.

Nature of Primitive Hexapoda.—Two divergent views have been held as to the nature of the original hexapod stock. Some of those zoologists who look to Peripatus, or a similar worm-like form, as representing the direct ancestors of the Hexapoda have laid stress on a larva like the caterpillar of a moth or saw-fly as representing a primitive stage. On the other hand, the view of F. Müller and F. Brauer, that the Thysanura represent more nearly than any other existing insects the ancestors of the class, has been accepted by the great majority of students. And there can be little doubt that this belief is justified. The caterpillar, or the maggot, is a specialized larval form characteristic of the most highly developed orders, while the campodeiform larva is the starting-point for the more primitive insects. The occurrence in the hypermetamorphic Coleoptera (see supra) of a campodeiform preceding an eruciform stage in the life-history is most suggestive. Taken in connexion with the likeness of the young among the more generalized orders to the adults, it indicates clearly a thysanuroid starting-point for the evolution of the hexapod orders. And we must infer further that the specialization of the higher orders has been accompanied by an increase in the extent of the metamorphosis—a very exceptional condition among animals generally, as has been ably pointed out by L. C. Miall (1895).

Origin of Wings.—The post-embryonic growth of Hexapods with or without metamorphosis is accompanied in most cases by the acquisition of wings. These organs, thus acquired during the lifetime of the individual, must have been in some way acquired during the evolution of the class. Many students of the group, following Brauer, have regarded the Apterygota as representing the original wingless progenitors of the Pterygota, and the many primitive characters shown by the former group lend support to this view. On the other hand, it has been argued that the presence of wings in a vast majority of the Hexapoda suggests their presence in the ancestors of the whole class. It is most unlikely that wings have been acquired independently by various orders of Hexapoda, and if we regard the Thysanura as the slightly modified representatives of a primitively wingless stock, we must postulate the acquisition of wings by some early offshoot of that stock, an offshoot whence the whole group of the Pterygota took its rise. How wings were acquired by these primitive Pterygota must remain for the present a subject for speculation. Insect wings are specialized outgrowths of certain thoracic segments, and are quite unrepresented in any other class of Arthropods. They are not, therefore, like the wings of birds, modified from some pre-existing structures (the fore-limbs) common to their phylum; they are new and peculiar structures. Comparison of the tracheated wings with the paired tracheated outgrowths on the abdominal segments of the aquatic campodeiform larva of may-flies (see fig. 27) led C. Gegenbaur to the brilliant suggestion that wings might be regarded as specialized and transformed gills. But a survey of the Hexapoda as a whole, and especially a comparative study of the tracheal system, can hardly leave room for doubt that this system is primitively adapted for atmospheric breathing, and that the presence of tracheal gills in larvae must be regarded as a special adaptation for temporary aquatic life. The origin of insect wings remains, therefore, a mystery, deepened by the difficulty of imagining any probable use for thoracic outgrowths, comparable to the wing-rudiments of the Exopterygota, in the early stages of their evolution.

Origin of Metamorphosis.—In connexion with the question whether metamorphosis has been gradually acquired, we have to consider two aspects, viz. the bionomic nature of metamorphosis, and to what extent it existed in primitive insects. Bionomically, metamorphosis may be defined as the sum of adaptations that have gradually fitted the larva (caterpillar or maggot) for one kind of life, the fly for another. So that we may conclude that the factors of evolution would favour its development. With regard to its occurrence in primitive insects, our knowledge of the geological record is most imperfect, but so far as it goes it supports the conclusion that holometabolism (i.e. extreme metamorphosis) is a comparatively recent phenomenon of insect life. None of the groups of existing Endopterygota have been traced with certainty farther back than the Mesozoic epoch, and all the numerous Palaeozoic insect-fossils seem to belong to forms that possessed only imperfect metamorphosis. The only doubt arises from the existence of insect remains, referred to the order Coleoptera, in the Silesian Culm of Steinkunzendorf near Reichenbach. The oldest larva known, Mormolucoides articulatus, is from the New Red Sandstone of Connecticut; it belongs to the Sialidae, one of the lowest forms of Holometabola. It is now, in fact, generally admitted that metamorphosis has been acquired comparatively recently, and Scudder in his review of the earliest fossil insects states that “their metamorphoses were simple and incomplete, the young leaving the egg with the form of the parent, but without wings, the assumption of which required no quiescent stage before maturity.”

It has been previously remarked that the phenomena of holometabolism are connected with the development of wings inside the body (except in the case of the fleas, where there are no wings in the perfect insect). Of existing insects 90% belong to the Endopterygota. At the same time we have no evidence that any Endopterygota existed amongst Palaeozoic insects, so that the phenomena of endopterygotism are comparatively recent, and we are led to infer that the Endopterygota owe their origin to the older Exopterygota. In Endopterygota the wings commence their development as invaginations of the hypodermis, while in Exopterygota the wings begin—and always remain—as external folds or evaginations. The two modes of growth are directly opposed, and at first sight it appears that this fact negatives the view that Endopterygota have been derived from Exopterygota.

Only three hypotheses as to the origin of Endopterygota can be suggested as possible, viz.:—(1) That some of the Palaeozoic insects, though we infer them to have been exopterygotous, were really endopterygotous, and were the actual ancestors of the existing Endopterygota; (2) that Endopterygota are not descended from Exopterygota, but were derived directly from ancestors that were never winged; (3) that the predominant division—i.e. Endopterygota—of insects of the present epoch are descended from the predominant—if not the sole—group that existed in the Palaeozoic epoch, viz. the Exopterygota. The first hypothesis is not negatived by direct evidence, for we do not actually know the ontogeny of any of the Palaeozoic insects; it is, however, rendered highly improbable by the modern views as to the nature and origin of wings in insects, and by the fact that the Endopterygota include none of the lower existing forms of insects. The second hypothesis—to the effect that Endopterygota are the descendants of apterous insects that had never possessed wings (i.e. the Apterygogenea of Brauer and others, though we prefer the shorter term Apterygota)—is rendered improbable from the fact that existing Apterygota are related to Exopterygota, not to Endopterygota, and by the knowledge that has been gained as to the morphology and development of wings, which suggest that—if we may so phrase it—were an apterygotous insect gradually to develop wings, it would be on the exopterygotous system. From all points of view it appears, therefore, probable that Endopterygota are descended from Exopterygota, and we are brought to the question as to the way in which this has occurred.

It is almost impossible to believe that any species of insect that has for a long period developed the wings outside the body could change this mode of growth suddenly for an internal mode of development of the organs in question, for, as we have already explained, the two modes of growth are directly opposed. The explanation has to be sought in another direction. Now there are many forms of Exopterygota in which the creatures are almost or quite destitute of wings. This phenomenon occurs among species found at high elevations, among others found in arid or desert regions, and in some cases in the female sex only, the male being winged and the female wingless. This last state is very frequent in Blattidae, which were amongst the most abundant of Palaeozoic insects. The wingless forms in question are always allied to winged forms, and there is every reason to believe that they have been really derived from winged forms. There are also insects (fleas, &c.) in which metamorphosis of a “complete” character exists, though the insects never develop wings. These cases render it highly probable that insects may in some circumstances become wingless, though their ancestors were winged. Such insects have been styled anapterygotous. In these facts we have one possible clue to the change from exopterygotism to endopterygotism, namely, by an intermediate period of anapterygotism.

Although we cannot yet define the conditions under which exopterygotous wings are suppressed or unusually developed, yet we know that such fluctuations occur. There are, in fact, existing forms of Exopterygota that are usually wingless, and that nevertheless appear in certain seasons or localities with wings. We are therefore entitled to assume that the suppressed wings of Exopterygota tend to reappear; and, speaking of the past, we may say that if after a period of suppression the wings began to reappear as hypodermal buds while a more rigid pressure was exerted by the cuticle, the growth of the buds would necessarily be inwards, and we should have incipient endopterygotism. The change that is required to transform Exopterygota into Endopterygota is merely that a cell of hypodermis should proliferate inwards instead of outwards, or that a minute hypodermal evaginated bud should be forced to the interior of the body by the pressure of a contracted cuticle.

If it should be objected that the wings so developed would be rudimentary, and that there would be nothing to encourage their development into perfect functional organs, we may remind the reader that we have already pointed out that imperfect wings of Exopterygota do, even at the present time under certain conditions, become perfect organs; and we may also add that there are, even among existing Endopterygota, species in which the wings are usually vestiges and yet sometimes become perfectly developed. In fact, almost every condition that is required for the change from exopterygotism to endopterygotism exists among the insects that surround us.

But it may perhaps be considered improbable that organs like the wings, having once been lost, should have been reacquired on the large scale suggested by the theory just put forward. If so, there is an alternative method by which the endopterygotous may have arisen from the exopterygotous condition. The sub-imago of the Ephemeroptera suggests that a moult, after the wings had become functional, was at one time general among the Hexapoda, and that the resting nymph of the Thysanoptera or the pupa of the Endopterygota represents a formerly active stage in the life-history. Further, although the wing-rudiments appear externally in an early instar of an exopterygotous insect, the earliest instars are wingless and wing-rudiments have been previously developing beneath the cuticle, growing however outwards, not inwards as in the larva of an endopterygote. The change from an exopterygote to an endopterygote development could, therefore, be brought about by the gradual postponement to a later and later instar of the appearance of the wing-rudiments outside the body, and their correlated growth inwards as imaginal disks. For in the post-embryonic development of the ancestors of the Endopterygota we may imagine two or three instars with wing-rudiments to have existed, the last represented by the sub-imago of the may-flies. As the life-conditions and feeding-habits of the larva and imago become constantly more divergent, the appearance of the wing-rudiments would be postponed to the pre-imaginal instar, and that instar would become predominantly passive.

Relationships of the Orders.—Reasons have been given for regarding the Thysanura as representing, more nearly than any other living group, the primitive stock of the Hexapoda. It is believed that insects of this group are represented among Silurian fossils. We may conclude, therefore, that they were preceded, in Cambrian times or earlier, by Arthropods possessing well developed appendages on all the trunk-segments. Of such Arthropods the living Symphyla—of which the delicate little Scutigerella is a fairly well-known example—give us some representation.

No indications beyond those furnished by comparative anatomy help us to unravel the phylogeny of the Collembola. In most respects, the shortened abdomen, for example, they are more specialized than the Thysanura, and most of the features in which they appear to be simple, such as the absence of a tracheal system and of compound eyes, can be explained as the result of degradation. In their insunken mouth and their jaws retracted within the head-capsule, the Collembola resemble the entotrophous division of the Thysanura (see Aptera), from which they are probably descended.

From the thysanuroid stock of the Apterygota, the Exopterygota took their rise. We have undoubted fossil evidence that winged insects lived in the Devonian and became numerous in the Carboniferous period. These ancient Exopterygota were synthetic in type, and included insects that may, with probability, be regarded as ancestral to most of the existing orders. It is hard to arrange the Exopterygota in a linear series, for some of the orders that are remarkably primitive in some respects are rather highly specialized in others. As regards wing-structure, the Isoptera with the two pairs closely similar are the most primitive of all winged insects; while in the paired mesodermal genital ducts, the elongate cerci and the conspicuous maxillulae of their larvae the Ephemeroptera retain notable ancestral characters. But the vestigial jaws, numerous Malpighian tubes, and specialized wings of may-flies forbid us to consider the order as on the whole primitive. So the Dermaptera, which retain distinct maxillulae and have no ectodermal genital ducts, have either specialized or aborted wings and a large number of Malpighian tubes. The Corrodentia retain vestigial maxillulae and two pairs of Malpighian tubes, but the wings are somewhat specialized in the Copeognatha and absent in the degraded and parasitic Mallophaga. The Plecoptera and Orthoptera agree in their numerous Malpighian tubes and in the development of a folding anal area in the hind-wing. As shown by the number and variety of species, the Orthoptera are the most dominant order of this group. Eminently terrestrial in habit, the differentiation of their fore-wings and hind-wings can be traced from Carboniferous, isopteroid ancestors through intermediate Mesozoic forms. The Plecoptera resemble the Ephemeroptera and Odonata in the aquatic habits of their larvae, and by the occasional presence of tufted thoracic gills in the imago exhibit an aquatic character unknown in any other winged insects. The Odonata are in many imaginal and larval characters highly specialized; yet they probably arose with the Ephemeroptera as a divergent offshoot of the same primitive isopteroid stock which developed more directly into the living Isoptera, Plecoptera, Dermaptera and Orthoptera.

All these orders agree in the possession of biting mandibles, while their second maxillae have the inner and outer lobes usually distinct. The Hemiptera, with their piercing mandibles and first maxillae and with their second maxillae fused to form a jointed beak, stand far apart from them. This order can be traced with certainty back to the early Jurassic epoch, while the Permian fossil Eugereon, and the living order—specially modified in many respects—of the Thysanoptera indicate steps by which the aberrant suctorial and piercing mouth of the Hemiptera may have been developed from the biting mouth of primitive Isopteroids, by the elongation of some parts and the suppression of others. The Anoplura may probably be regarded as a degraded offshoot of the Hemiptera.

The importance of great cardinal features of the life-history as indicative of relationship leads us to consider the Endopterygota as a natural assemblage of orders. The occurrence of weevils—among the most specialized of the Coleoptera—in Triassic rocks shows us that this great order of metabolous insects had become differentiated into its leading families at the dawn of the Mesozoic era, and that we must go far back into the Palaeozoic for the origin of the Endopterygota. In this view we are confirmed by the impossibility of deriving the Endopterygota from any living order of Exopterygota. We conclude, therefore, that the primitive stock of the former sub-class became early differentiated from that of the latter. So widely have most of the higher orders of the Hexapoda now diverged from each other, that it is exceedingly difficult in most cases to trace their relationships with any confidence. The Neuroptera, with their similar fore- and hind-wings and their campodeiform larvae, seem to stand nearest to the presumed isopteroid ancestry, but the imago and larva are often specialized. The campodeiform larvae of many Coleoptera are indeed far more primitive than the neuropteran larvae, and suggest to us that the Coleoptera—modified as their wing-structure has become—arose very early from the primitive metabolous stock. The antiquity of the Coleoptera is further shown by the great diversity of larval form and habit that has arisen in the order, and the proof afforded by the hypermetamorphic beetles that the campodeiform preceded the eruciform larva has already been emphasized.

In all the remaining orders of the Endopterygota the larva is eruciform or vermiform. The Mecaptera, with their predominantly longitudinal wing-nervuration, serve as a link between the Neuroptera and the Trichoptera, their retention of small cerci being an archaic character which stamps them as synthetic in type, but does not necessarily remove them from orders which agree with them in most points of structure but which have lost the cerci. The standing of the Trichoptera in a position almost ancestral to the Lepidoptera is one of the assured results of recent morphological study, the mobile mandibulate pupa and the imperfectly suctorial maxillae of the Trichoptera reappearing in the lowest families of the Lepidoptera. This latter order, which is not certainly known to have existed before Tertiary times, has become the most highly specialized of all insects in the structure of the pupa. Diptera of the sub-order Orthorrhapha occur in the Lias and Cyclorrhapha in the Kimmeridgian. The order must therefore be ancient, and as no evidence is forthcoming as to the mode of reduction of the hind-wings, nor as to the stages by which the suctorial mouth-organs became specialized, it is difficult to trace the exact relationship of the group, but the presence of cerci and a degree of correspondence in the nervuration of the fore-wings suggest the Mecaptera as possible allies. There seems no doubt that the suctorial mouth-organs of the Diptera have arisen quite independently from those of the Lepidoptera, for in the former order the sucker is formed from the second maxillae, in the latter from the first. The eruciform larva of the Orthorrhapha leads on to the headless vermiform maggot of the Cyclorrhapha, and in the latter sub-order we find metamorphosis carried to its extreme point, the muscid flies being the most highly specialized of all the Hexapoda as regards structure, while their maggots are the most degraded of all insect larvae. The Siphonaptera appear by the form of the larva and the nature of the metamorphosis to be akin to the Orthorrhapha—in which division they have indeed been included by many students. They differ from the Diptera, however, in the general presence of palps to both pairs of maxillae, and in the absence of a hypopharynx, so it is possible that their relationship to the Diptera is less close than has been supposed. The affinities of the Hymenoptera afford another problem of much difficulty. They differ from other Endopterygota in the multiplication of their Malpighian tubes, and from all other Hexapoda in the union of the first abdominal segment with the thorax. Specialized as they are in form, development and habit, they retain mandibles for biting, and in their lower sub-order—the Symphyta—the maxillae are hardly more modified than those of the Orthoptera. From the evidence of fossils it seems that the higher sub-order—Apocrita—can be traced back to the Lias, so that we believe the Hymenoptera to be more ancient than the Diptera, and far more ancient than the Lepidoptera. They afford an example—paralleled in other classes of the animal kingdom—of an order which, though specialized in some respects, retains many primitive characters, and has won its way to dominance rather by perfection of behaviour, and specially by the development of family life and helpful socialism, than by excessive elaboration of structure. We would trace the Hymenoptera back therefore to the primitive endopterygote stock. The specialization of form in the constricted abdomen and in the suctorial “tongue” that characterizes the higher families of the order is correlated with the habit of careful egg-laying and provision of food for the young. In some way it is assured among the highest of the Hexapoda—the Lepidoptera, Diptera and Hymenoptera—that the larva finds itself amid a rich food-supply. And thus perfection of structure and instinct in the imago has been accompanied by degradation in the larva, and by an increase in the extent of transformation and in the degree of reconstruction before and during the pupal stage. The fascinating difficulties presented to the student by the metamorphosis of the Hexapoda are to some extent explained, as he ponders over the evolution of the class.

Bibliography.—References to the older classical writings on the Hexapoda are given in the article on Entomology. At present about a thousand works and papers are published annually, and in this place it is possible to enumerate only a few of the most important among (mostly) recent memoirs that bear upon the Hexapoda generally. Further references will be found appended to the special articles on the orders (Aptera, Coleoptera, &c.).

General Works.—A. S. Packard, Text-book of Entomology (London, 1898); V. Graber, Die Insekten (Munich, 1877–1879); D. Sharp, Cambridge Natural History, vols. v., vi. (London, 1895–1899); L. C. Miall and A. Denny, Structure and Life-history of the Cockroach (London, 1886); B. T. Lowne, The Anatomy, Physiology, Morphology and Development of the Blow-fly (2 vols., London, 1890–1895); G. H. Carpenter, Insects: their Structure and Life (London, 1899); L. F. Henneguy, Les Insectes (Paris, 1904); J. W. Folsom, Entomology (New York and London, 1906); A. Berlese, Gli Insetti (Milan, 1906), &c. (Extensive bibliographies will be found in several of the above.)

Head and Appendages.—J. C. Savigny, Mémoires sur les animaux sans vertèbres (Paris, 1816); C. Janet, Essai sur la constitution morphologique de la tête de l’insecte (Paris, 1899); J. H. Comstock and C. Kochi (American Naturalist, xxxvi., 1902); V. L. Kellogg (ibid.); W. A. Riley (American Naturalist, xxxviii., 1904); F. Meinert (Entom. Tidsskr. i., 1880); H. J. Hansen (Zool. Anz. xvi., 1893); J. B. Smith (Trans. Amer. Phil. Soc. xix., 1896); H. Holmgren (Zeitsch. wiss. Zoolog. lxxvi., 1904); K. W. Verhoeff (Abhandl. K. Leop.-Carol. Akad. lxxxiv., 1905).

Thorax, Legs and Wings.—K. W. Verhoeff (Abhandl. K. Leop.-Carol. Akad. lxxxii., 1903); F. Voss (Zeits. wiss. Zool. lxxviii., 1905); F. Dahl (Arch. f. Naturgesch. 1, 1884); J. Demoor (Arch. de biol. x., 1890); J. Redtenbacher (Ann. Kais. naturhist. Museum, Wien, i., 1886); R. von Lendenfeld (S. B. Akad. Wissens., Wien, lxxxiii., 1881); J. H. Comstock and J. G. Needham (Amer. Nat., xxxii., xxxiii., 1898–1899); C. W. Woodworth (Univ. California Entom. Bull. i., 1906).

Abdomen and Appendages.—E. Haase (Morph. Jahrb. xv., 1889); R. Heymons (Morph. Jahrb. xxiv., 1896; Abhandl. K. Leop.-Carol. Akad. lxxiv., 1899); K. W. Verhoeff (Zool. Anz. xix., xx., 1896–1897); S. A. Peytoureau, Contribution à l’étude de la morphologie de l’armure génitale des insectes (Bordeaux, 1895); H. Dewitz (Zeits. wiss. Zool. xxv., xxviii., 1874, 1877); E. Zander (ibid. lxvi., lxvii., 1899–1900).

Nervous System.—H. Viallanes (Ann. Sci. Nat. Zool. [6], xvii., xviii., xix., [7] ii., iv., 1884–1887); S. J. Hickson (Quart. Journ. Micr. Sci. xxv., 1885); W. Patten (Journ. Morph. i., ii., 1887–1888); F. Plateau (Mém. Acad. Belg. xliii., 1888); V. Graber (Arch. mikr. Anat. xx., xxi., 1882).

Respiratory System.—J. A. Palmén, Zur Morphologie des Tracheensystems (Leipzig, 1877); F. Plateau (Mém. Acad. Belg. xiv., 1884); L. C. Miall, Natural History of Aquatic Insects (London, 1895).

Digestive System, &c.—L. Dufour (Ann. Sci. Nat., 1824–1860); V. Faussek (Zeits. wiss. Zool. xlv., 1887).

Malpighian Tubes.—E. Schindler (Zeits. wiss. Zool. xxx., 1878); W. M. Wheeler (Psyche vi., 1893); L. Cuénot (Arch. de biol. xiv., 1895).

Reproductive Organs.—H. V. Wielowiejski (Zool. Anz. ix., 1886); J. A. Palmén, Über paarige Ausführungsgänge der Geschlechtsorgane bei Insekten (Helsingfors, 1884); H. Henking (Zeits. wiss. Zool. xlix., li., liv., 1890–1892); F. Leydig (Zool. Jahrb. Anat. iii., 1889).

Embryology.—F. Blochmann (Morph. Jahrb. xii., 1887); A. Kovalevsky (Mém. Acad. St-Pétersbourg, xvi., 1871; Zeits. wiss. Zool. xlv., 1887); V. Graber (Denksch. Akad. Wissens., Wien, lvi., 1889); K. Heider, Die Embryonalentwicklung von Hydrophilus piceus (Jena, 1889); W. M. Wheeler (Journ. Morph. iii., viii., 1889–1893); E. Korschelt and K. Heider, Handbook of the Comparative Embryology of Invertebrates (trans. M. Bernard), (vol. iii., London, 1899); R. Heymons, Die Embryonalentwicklung von Dermapteren und Orthopteren (Jena, 1895) (also Zeits. wiss. Zool. liii., 1891, lxii., 1897; Anhang zu den Abhandl. K. Akad. d. Wissens., Berlin, 1896); A. Lécaillon (Arch. d’anat. micr. ii., 1898); J. Carrière and O. Burger (Abhandl. K. Leop.-Carol. Akad. lxix., 1897); K. Escherich (ibid. lxxvii., 1901); F. Schwangart (Zeits. wiss. Zool. lxxvi., 1904); R. Ritter (ib. li., 1890); E. Metchnikoff (ib. xvi., 1866); H. Uzel (Zool. Anz. xx., 1897); J. W. Folsom (Bull. Mus. Comp. Zool. Harvard., xxxvi., 1900).

Parthenogenesis and Paedogenesis.—T. H. Huxley (Trans. Linn. Soc. xxii., 1858); R. Leuckart, Zur Kenntnis des Generationswechsels und der Parthogenesis bei den Insekten (Frankfurt, 1858); N. Wagner (Zeits. wiss. Zool. xv., 1865); L. F. Henneguy (Bull. Soc. Philomath. [9], i. 1899); A. Petrunkevich (Zool. Jahrb. Anat. xiv., xvii., 1901–1903); P. Marchal (Arch. zool. exp. et gén. [4], ii., 1904); L. Doncaster (Quart. Journ. Micr. Sci. xlix., li., 1906–1907).

Growth and Metamorphosis.—A. Weismann (Zeits. wiss. Zool. xiii., xiv., 1863–1864); F. Brauer (Verh. zool.-bot. Gesellsch., Wien, xix., 1869); Sir J. Lubbock (Lord Avebury), Origin and Metamorphosis of Insects (London, 1874); L. C. Miall (Nature, liii., 1895); L. C. Miall and A. R. Hammond, Structure and Life-history of the Harlequin-fly (Oxford, 1900); J. Gonin (Bull. Soc. Vaud. Sci. Nat. xxx., 1894); C. de Bruyne (Arch. de biol. xv. (1898); D. Sharp (Proc. Inter. Zool. Congress, 1898); E. B. Poulton (Trans. Linn. Soc. v., 1891); T. A. Chapman (Trans. Ent. Soc., 1893).

Classification.—F. Brauer (S. B. Akad. Wiss., Wien, xci., 1885); A. S. Packard (Amer. Nat. xx.; 1886); C. Börner, A. Handlirsch, F. Klapalek (Zool. Anz. xxvii., 1904); G. Enderlein (Zool. Anz. xxvi., 1903).

Palaeontology.—S. H. Scudder, in Zittel’s Palaeontology (French trans., vol. ii., Paris, 1887, and Eng. trans., vol. i., London, 1900); C. Brongniart, Insectes fossiles des temps primaires (St-Étienne, 1894); A. Handlirsch, Die fossilen Insekten und die Phylogenie der rezenten Formen (Leipzig, 1906).

Phylogeny.—Brauer, Lubbock, Sharp, Börner, &c. (opp. cit.); P. Mayer (Jena, Zeits. Naturw. x., 1876); B. Grassi (Atti R. Accad. dei Lincei, Roma [4], iv., 1888, and Archiv ital. biol. xi., 1889); F. Müller, Facts and Arguments for Darwin (trans. W. S. Dallas, London, 1869); N. Zograf (Congr. Zool. Int., 1892); E. R. Lankester (Quart. Journ. Micr. Sci. xlvii., 1904); G. H. Carpenter (Proc. R. Irish Acad. xxiv., 1903; Quart. Journ. Micr. Sci. xlix., 1905).  (D. S.*; G. H. C.)