Plateau has made similar observations upon the respiration of spiders and scorpions; but, to his great surprise, he was unable, either by direct observation, or by the graphic method, or by projection, to discover the slightest respiratory movement of the exterior of the body. This can only be explained by supposing that inspiration and expiration in pulmonate Arachnida are “intrapulmonary,” and affect only the proper, respiratory organs. The fact is less surprising because of the wide zoölogical separation between Arachnida and insects.

g. The air-sacs

In flying insects the tracheæ are in certain parts of the body enlarged into sacs of various sizes. These air-sacs were first observed by Swammerdam in a beetle (Geotrupes) and afterwards by Sir John Hunter in the bee, Sprengel subsequently discovering them in other insects. Those of the cockroach were described and illustrated in a very elaborate and detailed way by Straus-Dürckheim (Figs. 424 and 425). These vesicles are without tænidia. In the locust (M. femur-rubrum) there is a pair of very large vesicles in the prothorax (Fig. 396). The five pairs of large abdominal air-sacs arise, independently of the main tracheæ, directly from branches originating from the spiracles. All these large sacs are superficial, lying directly beneath the hypodermis, while the smaller ones are buried among the muscles. We have detected 53 of these vesicles in the head.

In the honey-bee (Fig. 426) and humble bee (Fig. 427) as well as the flies there are two enormous air-sacs at the base of the abdomen. In larval and wingless insects these sacs are entirely absent.

The use of the air-sacs.—It was supposed by Hunter as well as by Newport, and the view has been generally held, that the use of these sacs is to lighten the weight, i.e. lessen the specific gravity of the body during flight. It has, however, been suggested to us by A. A. Packard that this view from the standpoint of physics is incorrect. It is evident that the wings have to support just as much weight when the insect is flying, whether the tracheæ and vesicles are filled with air or not, the body of the insect during flight not being lightened by the air in the sacs. The use of these numerous sacs, some of them very spacious, is to afford a greater supply of air or oxygen than that contained in the air-tubes alone, and thus to afford a greater breathing capacity. The sacs are largest in dragon-flies, moths, flies, and bees, which are swift of flight. When we compare the active movements of these insects on the wing with those of a caterpillar or maggot, it will be seen that the far greater muscular exertions of the volant insect create a demand for a sudden and abundant supply of air to correspond to the increased rapidity of respiration; and the enlargements of the air-tubes, rapidly filled with air at each inspiration, render it possible to supply the demand.

The case is thus seen to be very different from that of those fishes which, having a swimming-bladder, can in the water change the specific gravity of their bodies. The case of insects is almost exactly paralleled by that of birds, where, as stated by Wiedersheim, the air-sacs appear to form integral parts of the respiratory apparatus: “a greater amount of air can by their means pass in and out during inspiration and expiration, especially through the larger bronchi, and consequently there is less necessity for the expansion of the lung parenchyma.” In other words, the supply of air in these sacs, as in insects, increases the breathing capacity of the bird during flight. Wiedersheim’s retention of the old idea that the specific gravity of the body is lessened (p. 262) seems, however, to be incorrect, as the weight of the bird’s body is not diminished by the air contained in the sacs.

There are two chief morphological tracheal systems: 1. The open or normal and primitive (holopneustic) type, and 2. The closed, or secondary and adaptive, i.e. apneustic, type. The open system is characterized by the presence of the stigmata. Through them the air directly enters into the tracheal tubes, whose delicate walls allow the exchange of gases in the blood. This type occurs in all sexually mature individuals, and also in the greater number of larvæ.

The closed or apneustic tracheal system is distinguished either by the want of stigmata, or, if present, they are not open, and do not function, so that the tracheæ cannot communicate with the air. In such cases the direct oxygenation of the blood is effected through the delicate integument, especially over the surface of the body in general, or in certain specialized places where the gill-like expansions of the skin are rich in tracheæ; such outgrowths, generally tubular or leaf-like, are called by Palmén tracheal gills.

This closed form of the tracheal system only occurs in the larval stage of aquatic or parasitic insects, as in the Plectoptera (Ephemeridæ), Perlidæ, Odonata, and Trichoptera, besides single genera of other orders, i.e. among Coleoptera, Gyrinus, Pelobius, Cnemidotus, and the young larva of Elmis; in the aquatic caterpillar of Paraponyx; in certain Diptera (Corethra, Chironomus, etc.), and some of the parasitic Hymenoptera (Microgaster).

Palmén has discovered that in the nymphs of Ephemeridæ, Perlidæ, Odonata, and the larvæ of most Trichoptera the tracheal branch (stigmatal branch) sent from the longitudinal trachea to where the thoracic stigmata would be situated if present, or where their vestiges only exist, are aborted, becoming simple solid cords not filled with air (Fig. 436, vf, and 447, f, funiculus or stigmatic cord). In the imago, however, they resume their function, connecting with the open functional stigmata. In Corethra, in its earliest stages, the entire tracheal system is, like the stigmatic branch, a system of solid cords and empty of air. (Palmén.)

Embryology shows that these stigmatal branches are well developed, and are formed at the same time as the stigmata. It was also shown by Dewitz, in a posthumous paper (1890), that in the young larval stage of the Odonata and Ephemeridæ the tracheal system is at first an open one, and in some of the families (Libellulidæ, Agrionidæ, and Ephemeridæ) thoracic stigmata are seen at a very early stage. From numerous experiments Dewitz concludes that in the young stages of Odonata and Ephemeridæ there is an open tracheal system; certainly in very young nymphs the thoracic spiracles allow the air to pass out. Fully grown nymphs of Æschnidæ, Libellulidæ, and Agrionidæ are capable not only of forcing the air out, but also, like the perfect insect, of inhaling it. Moreover, he proved that the gills of Ephemeridæ and Agrionidæ are not indispensable for the maintenance of life, as the insects can live without them, breathing either through the skin or by the rectum, or in both ways. It would seem that while in freshly hatched or very young larvæ of aquatic insects of different orders the skin is so delicate as to allow of dermal respiration, in after life, when the skin becomes thicker and denser, these expansions (gills), provided with a very thin and delicate skin, of a necessity grow out from the walls of the body.

It thus appears that the closure and total or partial abolition of the stigmata are in adaptation to aquatic life, and that such insects have descended from terrestrial air-breathing winged forms. This is an important argument against the view that the wings are modified tracheal gills.

In this connection may be noticed the closure of the 2d and 3d thoracic stigmata in holopneustic insects. We have found on laying open the body of a Sphinx larva that a large number of tracheal branches are seen to arise from the prothoracic and from the first pair of abdominal stigmata. Now between these points there are no spiracles or any external signs of them, there being in Lepidoptera no mesothoracic or metathoracic spiracles. Yet the main lateral trachea between the prothoracic and first abdominal segments deviates from its course and bends down to send off a small shrivelled stigmatal branch or cord to a place where, did a spiracle exist, we should look for it. In the larva of Platysamia cecropia, a similar vestigial stigmata branch is present.

In the larva of Corydalus, also, a trachea as large as the main longitudinal one takes its origin and passes directly under the main trachea. Now both tracheæ send a stigmatal branch opposite to where the mesothoracic stigma should be, if present, i.e. on the hind edge of the segment.

Verson, moreover, has found in the freshly hatched silkworm vestiges of meso- and metathoracic stigmata, each consisting of a circle of high hypodermal cells radially arranged around a common centre. The stigmatal branch is long, but shrivelled; its peritoneum is widened out into several berry-like saccules filled with cell-elements. In profile these rudimentary stigmata appear as a series of high hypodermal cells, which form the basis of a short blind tube.

After the second moult there begins a peculiar transformation of the rudimentary stigmata. The stigmatal branch connected with them sends off at various points thick tufts of capillary tracheæ which press against the base of the blind tube. Gradually lengthening, they form a fold which continues to increase in length. The numerous tufts of tracheal capillaries extend beyond the inner surface of the two layers of which the developing wing consists, the berry-like saccules are drawn into the wing and converted into more or less thick tubes, which finally form the “veins.” It is clear, therefore, says Verson, as Landois claimed, that the wings of Lepidoptera must be regarded as in the fullest sense organs of respiration. (Zool. Anz., 1890, p. 116.)

The number of pairs of stigmata varies, especially in maggots or larval Diptera, in adaptation to their varied modes of life. The larvæ of most flies (Muscidæ) have a pair of peculiarly shaped processes on the prothoracic segment bearing spiracular openings, and two anal spiracles, while in Ctenophora atrata L. only the anal pair are present. In the rat-tailed maggots (Eristalis) the long caudal process ends in two stigmata forming a respiratory tube, which can be thrust out of the water for the reception of air. In the larval mosquito (Fig. 433) and its ally, Mochlonyx, a short thick dorsal tube arises from the penultimate segment of the body, in which the two main tracheæ end, opening outward by a single spiracular aperture. Other dipterous larvæ (Simulium, Tanypus, and Ceratopogon) possess no spiracles, the tracheal system being a closed one.

The larvæ of most water beetles (Dyticidæ, Hydrophilidæ) possess but two spiracles, which, as in maggots, are situated at the end of the body. The aquatic larva of Amphizoa, according to Hubbard, breathes much as in the Dyticidæ, by means of two large valvular spiracles placed close together at the end of the body; “closed or rudimentary stigmata also occur on the mesothorax and on abdominal segments one to seven inclusive.”

Hubbard adds: “The larva of Pelobius is wholly aquatic and breathes by branchiæ, but the obsolete stigmata are indicated precisely as in Amphizoa, with the exception of the last pair, which in Amphizoa are open spiracles, but in Pelobius are suppressed; the terminal eight segments being prolonged in a swimming stylet.”

From a review of the distribution of spiracles, and their atrophy, partial or total, it will be seen that there are intermediate stages between the open (holopneustic) and closed (apneustic) systems. These, following Schiner, Brauer, and Palmén, may be defined thus:

1. Metapneustic type.—The larvæ possess only a single pair of open stigmata situated at the end of the body. (The dipterous Eristalis, Tipula, Culex, Ptychoptera, Bittacomorpha (Plate I.) with certain Tachinidæ, and in Coleoptera, the larvæ of Dyticus, and allies of Hydrophilus and Cyphon.)

2. Propneustic type.—The pupæ of Corethra, Culex, etc., in which only the most anterior pair of spiracles are open.

3. Amphipneustic type.—Larvæ with a pair of open spiracles situated at each end of the body, the intermediate spiracles being closed. (Most dipterous larvæ, Musca, after the first moult, Œstridæ, Asilidæ, and Syrphus.)

4. Peripneustic type; with prothoracic and abdominal spiracles, the mesothoracic pair atrophied or closed. (The larvæ of Neuroptera, Mecoptera, Trichoptera, Lepidoptera, of most Coleoptera,^[72] of most Diptera, and of most of the Hymenoptera.^[73])

These differences in the number of functional spiracles are in direct relation with the surroundings of the insects, the physical conditions of existence evidently determining the position of the active functional open spiracles and the closure of those useless to the organism.