This consideration, I fancy, suggests to us the most probable explanation of the honey-tubes and honey-dew. Creatures that eat so much and reproduce so fast as the aphides are rapidly sucking up juices all the time from the plant on which they fasten, and converting most of the nutriment so absorbed into material for fresh generations. That is how they swarm so fast over all our shrubs and flowers. But if there is any one kind of material in their food in excess of their needs, they would naturally have to secrete it by a special organ developed or enlarged for the purpose. I don't mean that the organ would or could be developed all at once, by a sudden effort, but that as the habit of fixing themselves upon plants and sucking their juices grew from generation to generation with these descendants of originally winged insects, an organ for permitting the waste product to exude must necessarily have grown side by side with it. Sugar seems to have been such a waste product, contained in the juices of the plant to an extent beyond what the aphides could assimilate or use up in the production of new broods; and this sugar is therefore secreted by special organs, the honey-tubes. One can readily imagine that it may at first have escaped in small quantities, and that two pores on their last segment but two may have been gradually specialised into regular secreting organs, perhaps under the peculiar agency of the ants, who have regularly appropriated so many kinds of aphides as miniature milch cows. So completely have some species of ants come to recognise their own proprietary interest in the persons of the aphides, that they provide them with fences and cow-sheds on the most approved human pattern. Sometimes they build up covered galleries to protect their tiny cattle; and these galleries lead from the nest to the place where the aphides are fixed, and completely enclose the little creatures from all chance of harm. If intruders try to attack the farmyard, the ants drive them away by biting and lacerating them. Sir John Lubbock, who has paid great attention to the mutual relations of ants and aphides, has even shown that various kinds of ants domesticate various species of aphis. The common brown garden-ant, one of the darkest skinned among our English races, 'devotes itself principally to aphides which frequent twigs and leaves'; especially, so far as I have myself observed, the bright green aphis of the rose, and the closely allied little black aphis of the broad bean. On the other hand a nearly related reddish ant pays attention chiefly to those aphides which live on the bark of trees, while the yellow meadow-ants, a far more subterranean species, keep flocks and herds of the like-minded aphides which feed upon the roots of herbs and grasses.