The thorax is covered by the first three rings behind the head, and on it are six legs, two on each segment.  The remainder of the caterpillar is abdominal and carries small pro-legs with which to help it cling to twigs and leaves, and the heavy anal props that support the vent.  By using these and several of the pro-legs immediately before them, the caterpillar can cling and erect the front part of the body so that it can strike from side to side when disturbed.  In the case of caterpillars that have a horn, as Celeus, or sets of them as Regalis, in this attitude they really appear quite formidable, and often I have seen them drive away small birds, while many people flee shrieking.

There are little tubes that carry air to the trachea, as caterpillars have no lungs and can live with a very small amount of air.

The skin may be rough, granulated, or soft and fine as silk, and in almost every instance of exquisite colour:  bluish green, greenish blue, wonderful yellows and from pale to deep wine red, many species having oblique touches of contrasting colours on the abdominal rings.  Others are marked with small projections of bright colours from which tufts of hair or bristles may grow.  In some, as Io, these bristles are charged with an irritating acid that will sting for an hour after coming in contact with the skin, but does no permanent injury.  On a few there are what seem to be small pockets of acid that can be ejected with a jerk, and on some a sort of filament that is supposed to distil a disagreeable odour.  As the caterpillar only uses these when disturbed, it is safe to presume that they are placed for defence, but as in the case of moths I doubt their efficacy.

Some lepidopterists have thought the sex of a moth could be regulated by the amount of food given the caterpillar; but with my numerous other doubts I include this.  It is all of a piece with any attempt at sex regulation.  I regard it as morally certain that sex goes back to the ovary and that the egg produced yields a male or female caterpillar in the beginning.  I am becoming convinced that caterpillars recognize sex in each other, basing the theory on the facts that in half a dozen instances I have found cocoons, spun only a few inches apart.  One pair brought to me as interwoven.  Two of these are shown in the following chapter.  In all cases a male and female emerged within a few minutes of each other and mated as soon as possible.  If a single pair of these cocoons ever had produced two of a kind, it would give rise to doubts.  When all of them proved to be male and female that paired, it seems to me to furnish conclusive evidence that the caterpillars knew what they were doing, and spun in the same place for the purpose of appearing together.