We have already pointed out that the wall of the air-tube, for a short distance from the spiracular orifice, has a tesselated instead of a spiral marking. In the thoracic spiracles the tesselated cells are grouped round regularly placed setæ (fig. 85 _I_). The chitinous cuticle within the opening is crowded with fine setæ, which are often arranged so as to form a fringe on one or both sides of the internal aperture. (_Supra_, p. 152.) _Mechanism of Respiration._ In animals with a complete circulation, aërated blood is diffused throughout the body by means of arteries and capillaries, which deliver it under pressure at all points. Such animals usually possess a special aërating chamber (lung or gill), where oxygen is made to combine with the hæmoglobin of the blood. It is otherwise with Insects. Their blood escapes into great lacunæ, where it stagnates, or flows and ebbs sluggishly, and a diffuse form of the internal organs becomes necessary for their free exposure to the nutritive fluid. The blood is not injected into the tissues, but they are bathed by it, and the compact kidney or salivary gland is represented in Insects by tubules, or a thin sheet of finely divided lobules. By a separate mechanism, air is carried along ramified passages to all the tissues. Every organ is its own lung. We must now consider in more detail how air is made to enter and leave the body of an Insect. The spiracles and the air-tubes have been described, but these are not furnished with any means of creating suction or pressure; and the tubes themselves, though highly elastic, are non-contractile, and must be distended or emptied by some external force. Many Insects, especially such as fly rapidly, exhibit rhythmical movements of the abdomen. There is an alternate contraction and dilatation, which may be supposed to be as capable of setting up expirations and inspirations as the rise and fall of the diaphragm of a Mammal. In many Insects, two sets of muscles serve to contract the abdomen--viz., muscles which compress or flatten, and muscles which approximate or telescope the segments.[152] In the Cockroach the second set is feebly developed, but the first is more powerful, and causes the terga and sterna alternately to approach and separate with a slow, rhythmical movement; in a Dragon-fly or Humble-bee the action is much more conspicuous, and it is easy to see that the abdomen is bent as well as depressed at each contraction. No special muscles exist for dilating the abdomen, and this seems to depend entirely upon the elasticity of the parts. It was long supposed that, when the abdomen contracted, air was expelled from the body, and the air passages emptied; that when the abdomen expanded again by its own elasticity, the air passages were refilled, and that no other mechanism was needed. Landois pointed out, however, that this was not enough. Air must be forced into the furthest recesses of the tracheal system, where the