I am lingering over these details to show the importance which the good construction of her palisade of threads possesses for the Great Peacock. If ill-ordered, entangled and therefore intractable when pushed, the series of boxed cones will offer an insurmountable resistance and the Moth will perish, a victim of the caterpillar’s imperfect art. If constructed with mathematical accuracy, but with sparse rows of threads in insufficient numbers, it will leave the retreat exposed to dangers from without and the chrysalis will become the prey of some intruder, of whom there are many in search of somnolent nymphs, forming easy victims. For the caterpillar, therefore, this double-acting mouth is a work of the highest importance. It has to expend upon it all that it possesses in foresight, in gleams of reason and in art capable of modification when circumstances require; it must in short give proof of the best of which its talents are capable. Let us follow it in its labours; let us interpose the experimental test; and we shall learn some curious facts. The cocoon and its opening are constructed simultaneously. When it has woven this or that part of the general wall, the caterpillar turns about, if need be, and with its unbroken thread proceeds to continue the palisade of converging filaments. To this end it pokes its head to the end of the roughly-defined funnel and then withdraws it, doubling the thread as it goes. This alternation of thrusts and withdrawals results in a circle of doubled filaments, which do not adhere to one another. The shift is not a long one; when the palisade is a row the richer, the caterpillar resumes its work upon the shell, a task which it again abandons to busy itself with the funnel; and so on, over and over again, the emission of the gummy product being suspended when the threads are to be left free and copiously effected when they have to be stuck together in order to obtain a solid texture. The exit-funnel is not, as we see, a piece of work executed continuously; the caterpillar works at it intermittently, as the general shell progresses. From the beginning to the end of its spinning-period, so long as the reservoirs of silk are not exhausted, it multiplies the tiers without neglecting the rest of the cocoon. These tiers take the form of cones enclosed one within the other and of increasingly obtuse angles, until the last to be spun are so flat as to become almost level surfaces. If nothing happen to disturb the worker, the work is performed with a perfection that would do credit to a discerning industry capable of realizing the why and wherefore of things. Can the caterpillar be said to have any conception, however slight, of the importance of its task, of the future function of its overlapping conical palisades? This is what we are about to learn.