But the spinning mammillae or spinnerets are still more characteristic and more easily seen, though, curiously enough, it is not among the cleverest spinners that they are most conspicuous. In the family to which most of the cellar spiders belong (Agelenidae) and in the elongate brown or mouse-coloured spiders found lurking under stones (Drassidae) they are visible as little finger-like projections at the posterior end of the abdomen, but if we have taken our specimen from a circular web (Epeiridae) we shall have to look for them more closely. In these spiders they are beneath the abdomen near its termination, and are not visible from above. Moreover when at rest their tips are applied together so that they form a small rosette in surface-view, or, in profile, a slight cone. The best way to capture a spider for examination is to induce it to run up into a small glass specimen tube—for spiders readily part with their legs if handled roughly—and if we have adopted this method we shall see the spinnerets in use as the animal crawls about the tube. It will not move without first attaching a silken cable to the glass, and this cable lengthens as the spider progresses, so that before long the interior of the tube will be a network of silken threads, and its sides will be flecked with little white specks where the threads have been re-attached for a new departure; and by observing closely we shall be able to note the extreme mobility of the spinnerets in action. All spiders spin, but it is by no means all spiders that make snares for the purpose of catching prey. The fundamental purpose of the spinning organs seems to be to connect the spider with its point of departure. The jumping spiders (Attidae) make no snare, but this “drag-line” as it has been called comes in very useful when stalking prey on the vertical surface of a wall, when a miscalculation at the moment of pouncing upon it would entail a considerable fall were it not for such an anchorage. It can hardly be doubted—though of course it is incapable of proof—that all the more complicated spinning operations originated in this universal spider habit, but all known spiders have learnt to apply their power of making silk to other purposes. If they do not make snares they at least spin “cocoons” for the protection of their eggs, and if they have a definite home from which they emerge to seek food, such a retreat is always more or less lined with silk. It is clear that a spider cocoon is quite different from that of an insect; it encloses the eggs and is manufactured by the mother, whereas among the insects the larva makes the cocoon for the protection of the pupa or chrysalis into which it is about to turn. However far from exhaustive the foregoing study of spider structure may be it will suffice for our purposes, at least for the present, and we may proceed at once to an investigation of one of the most remarkable achievements