Sternal Muscles of Thorax.—Two tubular apodemes, lying one behind the other, project into the thorax from the ventral surface (p. 59 and fig. 27). To the foremost of these are attached three paired muscles and one median muscle. The median muscle passes to the second tubular apodeme. The anterior pair pass forwards and outwards to the base of the prothoracic leg; the next pair directly outwards to the base of the middle leg; while the posterior pair pass outwards and backwards to the arms of the medifurca. From the second tubular apodeme, in front of the metasternum, four pairs of muscles spring. Those of the anterior pass forwards and outwards to the coxa of the fore limb; the second pair directly outwards to the base of the metathoracic legs; the third pair backwards and outwards to the arms of the postfurca; the fourth pair backwards to the second abdominal sternum.

The muscles attached to the medi- and postfurca (other than those connecting them with the tubular apodemes) are:— (1) A pair passing from the posterior edge of the arms of the medifurca to the stem of the postfurca; (2) a pair which diverge from the stem of the postfurca and proceed to the fore part of the second abdominal sternum; (3) a pair passing from the posterior edge of the arms of the postfurca, these are directed inwards and backwards, and are inserted into the hinder part of the second abdominal sternum; (4) a pair already mentioned, which correspond in position and action to the tergo-sternal muscles, and spring from the stem of the postfurca, passing upwards and outwards to the sides of the first abdominal tergum.

The muscles attached to the arms of each furca pass to other structures in or near the middle line of the body. The pull of such muscles must alter the slope of the two steps in the ventral floor of the thorax (p. 58, and fig. 3, p. 12). When the furca is drawn forwards, the step is rendered vertical or even inclined forward, the sterna being approximated; while, on the other hand, a backward pull brings the step into a horizontal position, and separates the sterna.

Tergal Muscles of Thorax.—The longitudinal tergal muscles are much reduced in width when compared with those of the abdomen. Sets of obliquely placed muscles, which may be called the lateral thoracic muscles, arise from near the middle of each tergum, and converge to tendinous insertions on the fore edge of each succeeding tergum, close to the lateral wall of the body.

The principal muscles of the legs are figured and named, and their action can readily be inferred from the names assigned to them.

Insect Mechanics.

The mechanics of Insect movements require exposition and illustration far beyond what is possible in a book like this. Even the elaborate dissections of Lyonnet and Straus-Dürckheim are not a sufficient basis for a thorough treatment of the subject, and until we possess many careful dissections, made by anatomists who are bent upon mastering the action of the parts, our views must needs be vague and of doubtful value. Zoologists of great eminence have been led into erroneous statements when they have attempted to characterise shortly a complex animal mechanism which they did not think it worth while to analyse completely.86

The action of flight and the muscles attached to the wings are best studied in Insects of powerful flight. The female Cockroach cannot fly at all, and the male is by no means a good flier. Both sexes are, however, admirably fitted for running.

In running, two sets, each consisting of three legs, move simultaneously. A set includes a fore and hind limb of the same side and the opposite middle leg. Numbering them from before backwards, and distinguishing the right and left sides by their initial letters, we can represent the legs which work together as—

R₁   L₂   R₃


L₁   R₂   L₃

The force exerted by Insects has long been remarked with surprise, and it is a fact familiar, not only to naturalists, but to all observant persons, that, making allowance for their small size, Insects are the most powerful of common animals. Popular books of natural history give striking and sometimes exaggerated accounts of the prodigious strength put forth by captive Insects in their efforts to escape. Thus we are told that the flea can draw 70 or 80 times its own weight.87 The Cockchafer is said to be six times as strong as a horse, making allowance for size. A caterpillar of the Goat Moth, imprisoned beneath a bell-glass, weighing half a pound, which was loaded with a book weighing four pounds, nevertheless raised the glass and made its escape.

This interesting subject has been investigated by Plateau,88 who devised the following experiment. The Insect to be tested was confined within a narrow horizontal channel, which was laid with cloth. A thread attached to its body was passed over a light pulley, and fastened to a small pan, into which sand was poured until the Insect could no longer raise it. Some of the results are given in the following table:—

Table of Relative Muscular Force of Insects (Plateau).

One obvious result is that within the class of Insects the relative muscular force (as commonly understood) is approximately in the inverse proportion of the weight—that is, the strength of the Insect is (by this mode of calculation) most conspicuous in the smaller species.

In a later memoir89 Plateau gives examples from different Vertebrate and Invertebrate animals, which lead to the same general conclusion.

Ratio of weight drawn to weight of body (Plateau).

The inference commonly drawn from such data is that the muscles of small animals possess a force which greatly exceeds that of large quadrupeds or man, allowance being made for size, and that the explanation of this superior force is to be looked for in some peculiarity of composition or texture. Gerstaecker,90 for example, suggests that the higher muscular force of Arthropoda may be due to the tender and yielding nature of their muscles. An explanation so desperate as this may well lead us to inquire whether we have understood the facts aright. Plateau’s figures give us the ratio of the weight drawn or raised to the weight of the animal. This we may, with him, take as a measure of the relative muscular force. In reality, it is a datum of very little physiological value. By general reasoning of a quite simple kind it can be shown that, for muscles possessing the same physical properties, the relative muscular force necessarily increases very rapidly as the size of the animal decreases. For the contractile force of muscles of the same kind depends simply upon the number and thickness of the fibres, i.e., upon the sectional area of the muscles. If the size of the animal and of its muscles be increased according to any uniform scale, the sectional area of a given muscle will increase as the square of any linear dimension. But the weight increases in a higher proportion, according to the increase in length, breadth, and depth jointly, or as the cube of any linear dimension.91 The ratio of contractile force to weight must therefore become rapidly smaller as the size of the animal increases. Plateau’s second table (see above) actually gives a value for the relative muscular force of the Bee, in comparison with the Horse, which is only one-fourteenth of what it ought to turn out, supposing that both animals were of similar construction, and that the muscular fibres in both were equal in contractile force per unit of sectional area.92

A later series of experiments93 brings out this difference in a precise form. Plateau has determined by ingenious methods what he calls the Absolute Muscular Force 94 of a number of Invertebrate animals (Lamellibranch Mollusca, and Crustacea), comparing them with man and other Vertebrates. His general conclusions may be shortly given as follows:—The absolute muscular force of the muscles closing the pincers of Crabs is low in comparison with that of Vertebrate muscles. The absolute force of the adductor muscles closing a bivalve shell may, in certain Lamellibranchs, equal that of the most powerful Mammalian muscles; in others it falls below that of the least powerful muscles of the frog, which are greatly inferior in contractile force to Mammalian muscles. We find, therefore, that the low contractile force of Insect muscles is in harmony, and not in contrast, with common observation of their physical properties, and that the high relative muscular force, correctly enough attributed to them, is explicable by considerations which apply equally well to models or other artificial structures.

The comparison between the muscular force of Insects and large animals is sometimes made in another way. For example, in Carpenter’s Zoology95 the spring of the Cheese-hopper is described, and we are told that “the height of the leap is often from twenty to thirty times the length of the body; exhibiting an energy of motion which is particularly remarkable in the soft larva of an Insect. A Viper, if endowed with similar powers, would throw itself nearly a hundred feet from the ground.” It is here implied that the equation

Height of Insect’s leapLength of Insect =  Supposed ht. of Viper’s leap (100 ft.)Length of Viper

should hold if the two animals were “endowed with similar powers.”

But it is known that the work done by contraction of muscles of the same kind is proportional to the volume of the muscles (“Borelli’s Law”),96 and in similar animals the muscular volumes are as the weights. Therefore the equation

Work of InsectWeight of Insect =  Work of ViperWeight of Viper

will more truly represent the imaginary case of equal leaping power. But the work = weight raised × height, and the weight raised is in both cases the weight of the animal itself. Therefore

Wt. × Ht.Wt. (Insect) =  Wt. × Ht.Wt. (Viper),

and Ht. (Insect) = Ht. (Viper). The Viper’s efficiency as a leaping animal would, therefore, equal that of a Cheese-hopper if it leaped the same vertical height. Therefore, if the two animals were “endowed with similar powers,” the heights to which they could leap would be equal, and not proportional to their lengths, as is assumed in the passage quoted.

Straus-Dürckheim observes that a Flea can leap a foot high, which is 200 times its own length, and this has been considered a stupendous feat. It is really less remarkable than a schoolboy’s leap of two feet, for it indicates precisely as great efficiency of muscles and other leaping apparatus as would be implied in a man’s leap to the same height, viz., one foot.97

The Fat-body.

Adhering to the inner face of the abdominal wall is a cellular mass, which forms an irregular sheet of dense white appearance. This is the fat-body. Its component cells are polygonal, and crowded together. When young they exhibit nuclei and vacuolated protoplasm, but as they get older the nuclei disappear, the cell-boundaries become indistinct, and a fluid, loaded with minute refractive granules,98 takes the place of the living protoplasm. Rhombohedral or hexagonal crystals, containing uric acid, form in the cells and become plentiful in old tissue. The salt (probably urate of soda) is formed by the waste of the proteids of the body. What becomes of it in the end we do not know for certain, but conjecture that it escapes by the blood which bathes the perivisceral cavity, that it is taken up again by the Malpighian tubules, and is finally discharged into the intestine. The old gorged cells probably burst from time to time, and the infrequency of small cells among them renders it probable that rejuvenescence takes place, the burst cells passing through a resting-stage, accompanied by renewal of their nuclei, and then repeating the cycle of change.

The segmental tubes forming the Wolffian body of Vertebrates have at first no outlet, and embryologists have hesitated to regard this phase of development as the permanent condition of any ancestral form.99 It is, therefore, of interest to find in the fat-body of the Cockroach an example of a solid, mesoblastic, excretory organ, functional throughout life, but without efferent duct.

The fat-body is eminently a metabolic tissue, the seat of active chemical change in the materials brought by the blood. Its respiratory needs are attested by the abundant air-tubes which spread through it in all directions.

The considerable bulk of the fat-body in the adult Cockroach points to the unusual duration of the perfect Insect. It is usually copious in full-fed larvæ, but becomes used up in the pupa-stage.

Extensions of the fat-body surround the nervous chain, the reproductive organs and other viscera. Sheets of the same substance lie in the pericardial sinus on each side of the heart.

The Cœlom.

The fat-body is in reality, as development shows, the irregular cellular wall of the cœlom, or perivisceral space. Through this space courses the blood, flowing in no defined vessels, but bathing all the walls and viscera. In other words, the fat-body is an aggregation of little-altered mesoblast-cells, excavated by the cœlom, the rest of the mesoblast having gone to form the muscular layers of the body-wall and of the digestive tube.

CHAPTER VI.

The nervous system of the Cockroach comprises ganglia and connectives,100 which extend throughout the body. We have first, a supra-œsophageal ganglion, or brain, a sub-œsophageal ganglion, and connectives which complete the œsophageal ring. All these lie in the head; behind them, and extending through the thorax and abdomen, is a gangliated cord, with double connectives. The normal arrangement of the ganglia in Annulosa, one to each somite, becomes more or less modified in Insects by coalescence or suppression, and we find only eleven ganglia in the Cockroach—viz., two cephalic, three thoracic, and six abdominal.

The nervous centres of the head form a thick, irregular ring, which swells above and below into ganglionic enlargements, and leaves only a small central opening, occupied by the œsophagus. The tentorium separates the brain or supra-œsophageal ganglion from the sub-œsophageal, while the connectives traverse its central plate. Since the œsophagus passes above the plate, the investing nervous ring also lies almost wholly above the tentorium.

The brain is small in comparison with the whole head; it consists of two rounded lateral masses or hemispheres, incompletely divided by a deep and narrow median fissure. Large optic nerves are given off laterally from the upper part of each hemisphere; lower down, and on the front of the brain, are the two gently rounded antennary lobes, from each of which proceeds an antennary nerve; while from the front and upper part of each hemisphere a small nerve passes to the so-called “ocellus,” a transparent spot lying internal to the antennary socket on each side in the suture between the clypeus and the epicranium. The sub-œsophageal ganglion gives off branches to the mandibles, maxillæ, and labrum. While, therefore, the supra-œsophageal is largely sensory, the sub-œsophageal ganglion is the masticatory centre.

The œsophageal ring is double below, being completed by the connectives and the sub-œsophageal ganglion; also by a smaller transverse commissure, which unites the connectives, and applies itself closely to the under-surface of the œsophagus.101

Two long connectives issue from the top of the sub-œsophageal ganglion, and pass between the tentorium and the submentum on their way to the neck and thorax. The three thoracic ganglia are large (in correspondence with the important appendages of this part of the body) and united by double connectives. The six abdominal ganglia have also double connectives, which are bent in the male, as if to avoid stretching during forcible elongation of the abdomen. The sixth abdominal ganglion is larger than the rest, and is no doubt a complex, representing several coalesced posterior ganglia; it supplies large branches to the reproductive organs, rectum, and cerci.

Internal Structure of Ganglia.

Microscopic examination of the internal structure of the nerve-cord reveals a complex arrangement of cells and fibres. The connectives consist almost entirely of nerve-fibres, which, as in Invertebrates generally, are non-medullated. The ganglia include (1) rounded, often multipolar, nerve-cells; (2) tortuous and extremely delicate fibres collected into intricate skeins; (3) commissural fibres, and (4) connectives. The chief fibrous tracts are internal, the cellular masses outside them. A relatively thick, and very distinct neurilemma, probably chitinous, encloses the cord. Its cellular matrix, or chitinogenous layer, is distinguished by the elongate nuclei of its constituent cells.102 Tracheal trunks pass to each ganglion, and break up upon and within it into a multitude of fine branches.

Bundles of commissural fibres pass from the ganglion cells of one side of the cord to the peripheral nerves of the other. There are also longitudinal bands which blend to form the connectives, and send bundles into the peripheral nerves. Of the peripheral fibres, some are believed to pass direct to their place of distribution, while others traverse at least one complete segment and the corresponding ganglion before separating from the cord.

Median Nerve-Cord.

Lyonnet,103 Newport,104 and Leydig105 have found in large Insects a system of median nerves, named respiratory (Newport) or sympathetic (Leydig). These nerves do not form a continuous cord extending throughout the body, but take fresh origin in each segment from the right and left longitudinal commissures alternately. The median nerve lies towards the dorsal side of the principal nerve-cord, crosses over the ganglion next behind, and receives a small branch from it. Close behind the ganglion it bifurcates, the branches passing outwards and blending with the peripheral nerves. Each branch, close to its origin, swells into a ganglionic enlargement. The median nerve and its branches differ in appearance and texture from ordinary peripheral nerves, being more transparent, delicate, and colourless. They are said to supply the occlusor muscles of the stigmata. In the Cockroach the median nerves are so slightly developed in the thorax and abdomen (if they actually exist) that they are hardly discoverable by ordinary dissection. We have found only obscure and doubtful traces of them, and these only in one part of the abdominal nerve-cord. The stomato-gastric nerves next to be described appear to constitute a peculiar modification of that median nerve-cord which springs from the circum-œsophageal connectives.

Stomato-gastric Nerves.

In the Cockroach the stomato-gastric nerves found in so many of the higher Invertebrates are conspicuously developed. From the front of each œsophageal connective, a nerve passes forwards upon the œsophagus, outside the chitinous crura of the tentorium. Each nerve sends a branch downwards to the labrum, and the remaining fibres, collected into two bundles, join above the œsophagus to form a triangular enlargement, the frontal ganglion. From this ganglion a recurrent nerve passes backwards through the œsophageal ring, and ends on the dorsal surface of the crop (·3 inch from the ring), in a triangular ganglion, from which a nerve is given off outwards and backwards on either side. Each nerve bifurcates, and then breaks up into branches which are distributed to the crop and gizzard.106 Just behind the œsophageal ring, the recurrent nerve forms a plexus with a pair of nerves which proceed from the back of the brain. Each nerve forms two ganglia, one behind the other, and each ganglion sends a branch inwards to join the recurrent nerve. Fine branches proceed from the paired nerves of the œsophageal plexus to the salivary glands.

The stomato-gastric nerves differ a good deal in different insects; Brandt107 considers that the paired and unpaired nerves are complementary to each other, the one being more elaborate, according as the other is less developed. A similar system is found in Mollusca, Crustacea, and some Vermes (e.g., Nemerteans). When highly developed, it contains unpaired ganglia and nerves, but may be represented only by an indefinite plexus (earthworm). It always joins the œsophageal ring, and sends branches to the œsophagus and fore-part of the alimentary canal. The system has been identified with the sympathetic, and also with the vagus of Vertebrates, but such correlations are hazardous; the first, indeed, may be considered as disproved.

Internal Structure of Brain.

The minute structure of the brain has been investigated by Leydig, Dietl, Flögel, and others, and exhibits an unexpected complexity. It is as yet impossible to reduce the many curious details which have been described to a completely intelligible account. The physiological significance, and the homologies of many parts are as yet altogether obscure. The comparative study of new types will, however, in time, bridge over the wide interval between the Insect-brain and the more familiar Vertebrate-brain, which is partially illuminated by physiological experiment. Mr. E. T. Newton has published a clear and useful description108 of the internal and external structure of the brain of the Cockroach, which incorporates what had previously been ascertained with the results of his own investigations. He has also described109 an ingenious method of combining a number of successive sections into a dissected model of the brain. Having had the advantage of comparing the model with the original sections, we offer a short abstract of Mr. Newton’s memoir as the best introduction to the subject. He describes the central framework of the Cockroach brain as consisting of two solid and largely fibrous trabeculæ, which lie side by side along the base of the brain, becoming smaller at their hinder ends; they meet in the middle line, but apparently without fusion or exchange of their fibres. Each trabecula is continued upwards by two fibrous columns, the cauliculus in front, and the peduncle behind; the latter carries a pair of cellular disks, the calices (the cauliculus, though closely applied to the calices, is not connected with them); these disks resemble two soft cakes pressed together above, and bent one inwards, and the other outwards below. The peduncle divides above, and each branch joins one of the calices of the same hemisphere.

This central framework is invested by cortical ganglionic cells, which possess distinct nuclei and nucleoli. A special cellular mass forms a cap to each pair of calices, and this consists of smaller cells without nucleoli. Above the meeting-place of the trabeculæ is a peculiar laminated mass, the central body, which consists of a network of fibres continuous with the neighbouring ganglionic cells, and enclosing a granular substance. The antennary lobes consist of a network of fine fibres enclosing ganglion cells, and surrounded by a layer of the same. It is remarkable that no fibrous communications can be made out between the calices and the cauliculi, or between the trabeculæ and the œsophageal connectives.

The sense organs of Insects are very variable, both in position and structure. Three special senses are indicated by transparent and refractive parts of the cuticle, by tense membranes with modified nerve-endings, and by peculiar sensory rods or filaments upon the antennæ. These are taken to be the organs respectively of sight, hearing, and smell. Other sense organs, not as yet fully elucidated, may co-exist with these. The maxillary palps of the Cockroach, for example, are continually used in exploring movements, and may assist the animal to select its food; the cerci, where these are well-developed, and the halteres of Diptera, have been also regarded as sense organs of some undetermined kind, but this is at present wholly conjecture.110

The compound eyes of the Cockroach occupy a large, irregularly oval space (see fig. 50) on each side of the head. The total number of facets may be estimated at about 1,800. The number is very variable in Insects, and may either greatly exceed that found in the Cockroach, or be reduced to a very small one indeed. According to Burmeister, the Coleopterous genus Mordella possesses more than 25,000 facets. Where the facets are very numerous, the compound eyes may occupy nearly the whole surface of the head, as in the House-fly Dragon-fly, or Gad-fly.

Together with compound eyes, many Insects are furnished also with simple eyes, usually three in number, and disposed in a triangle on the forehead. The white fenestræ, which in the Cockroach lie internal to the antennary sockets, may represent two simple eyes which have lost their dioptric apparatus. In many larvæ only simple eyes are found, and the compound eye is restricted to the adult form; in larval Cockroaches, however, the compound eye is large and functional.

Each facet of the compound eye is the outermost element of a series of parts, some dioptric and some sensory, which forms one of a mass of radiating rods or fibres. The facets are transparent, biconvex, and polygonal, often, but not quite regularly, hexagonal. In many Insects the deep layer of each facet is separable, and forms a concavo-convex layer of different texture from the superficial and biconvex lens. The facets, taken together, are often described as the cornea; they represent the chitinous cuticle of the integument. The subdivision of the cornea into two layers of slightly different texture suggests an achromatic correction, and it is quite possible, though unproved, that the two sets of prisms have different dispersive powers. Beneath the cornea we find a layer of crystalline cones, each of which rests by its base upon the inner surface of a facet, while its apex is directed inwards towards the brain. The crystalline cones are transparent, refractive, and coated with dark pigment; in the Cockroach they are comparatively short and blunt. Behind each cone is a nerve-rod (rhabdom), which, though outwardly single for the greater part of its length, is found on cross-section to consist of four components (rhabdomeres)111; these diverge in front, and receive the tip of a cone, which is wedged in between them; the nerve-rods are densely pigmented. The rhabdom is invested by a protoplasmic sheath, which is imperfectly separated into segments (retinulæ), corresponding in number with the rhabdomeres. Each retinula possesses at least one nucleus. The retinulæ were found by Leydig to possess a true visual purple. To the hinder ends of the retinulæ are attached the fibres of the optic nerve, which at this point emerges through a “fenestrated membrane.”