So far we have accounted for the spools, and for one pair of spigots—those on the anterior spinnerets. The lower spigot (_b_) on the middle spinneret often assists in laying down a foundation line when extra strength is required. In that case the line is fourfold, and can easily be split into four along its whole length, the threads from the middle spinnerets being rather finer than those from the anterior, but composed of the same kind of silk. There remain seven pairs of spigots whose function has still to be explained, two on the middle and five on the posterior spinnerets. The three which are clustered together on the posterior spinneret do not form _silk_ at all, that is, the material they emit does not harden on exposure to the air but remains fluid and adhesive. When the spider is spinning the “viscid spiral” of its web it is from these spigots that the sticky matter oozes, enveloping the true silken lines and presently resolving itself into little globules in the manner already described. The remaining spigots—two on the middle and two on the posterior spinnerets are employed only in spinning the egg-cocoon, and the silk they produce is unlike that used in making the snare, being much stronger and less elastic, and—in the case of the garden-spider—of a yellow colour. In the occasional attempts which have been made to substitute spiders for silkworms as commercial silk producers, it is only this cocoon silk that has given any considerable results, the produce of the other glands being far too frail for profitable use. Such attempts, however, have always failed, principally for a reason quite unconnected with the particular nature of the silk, namely, the difficulty of keeping the spiders in captivity. It is a simple matter to supply dozens of silkworms in the same box with mulberry leaves, but spiders require separate compartments or they will fight and devour each other, and the provision of suitable food for them is such a troublesome matter that it has proved quite impracticable on a commercial scale.