In order to understand the nature of the apparatus by which a caterpillar spins its silk, it is to be recollected that its whole interior structure differs from that of warm-blooded animals. It has, properly speaking, no heart, though a long tubular _dorsal vessel_, which runs along the back, and pulsates from twenty to one hundred times per minute, has been called so by Malpighi and others, but neither Lyonnet nor Cuvier could detect any vessel issuing from it, and consequently the fluid which is analogous to blood has no circulation. It differs also from the higher orders of animals in having no brain, the nerves running along the body being only united by little knobs, called ganglions. Another circumstance is, that it has no lungs, and does not breathe by the mouth, but by air-holes, or spiracles, eighteen in number, situated along the sides, in the middle of the rings, as may be seen in the following figure from Lyonnet. [Illustration: Caterpillar of the Goat Moth (_Cossus ligniperda_).] These spiracles communicate on each side with tubes, that have been called the wind-pipes (_tracheæ_). The spinning apparatus is placed near the mouth, and is connected with the silk-bags, which are long, slender, floating vessels, containing a liquid gum. The bags are closed at their lower extremity, become wider towards the middle, and more slender towards the head, where they unite to form the spinning-tube, or spinneret. The bags being in most cases longer than the body of the caterpillar, necessarily lie in a convoluted state, like the intestines of quadrupeds. The capacity, or rather the length, of the silk-bags is in proportion to the quantity of silk required for spinning; the _Cossus ligniperda_, for example, from living in the wood of trees spins little, having a bag only one-fourth the length of that of the silk-worm, though the caterpillar is at least twice the dimensions of the latter. The following figure, taken from the admirable treatise of Lyonnet on the anatomy of the _Cossus_, will render these several organs more easily understood than any description. The spinneret itself was supposed by Réaumur to have two outlets for the silk; but Lyonnet, upon minute dissection, found that the two tubes united into one before their termination; and he also assured himself that it was composed of alternate slips of horny and membranaceous substance,--the one for pressing the thread into a small diameter, and the other for enlarging it at the insect's pleasure. It is cut at the end somewhat like a writing-pen, though with less of a slope, and is admirably fitted for being applied to objects to which it may be required to attach silk. The following are magnified figures of the spinneret of the _Cossus_, from Lyonnet.