In vegetables, I grew three species of sweet potatoes— yellow, purple, and white skinned, and which differ also in their leaves and flowers; cabbages, kidney-beans, pumpkins, yuccas (Jatropha manihot), quequisque (a species of arum, Colocasia esculenta), lettuces, tomatoes, capsicums, endives, parsley, and carrots.

The climate was too damp to grow onions; neither could I succeed with peas, potatoes, or turnips. Scarlet runners (Phaseolus multiflorus) grew well, and flowered abundantly, but never produced a single pod. Darwin has shown that this flower is dependent, like many others, for its fertilisation upon the operations of the busy humble-bee, and that it is provided with a wonderful mechanism, by means of which its pollen is rubbed into the head of the bee, and received on the stigma of the next plant visited.* There are many humble-bees, of different species from ours, in tropical America; but none of them frequented the flowers of the scarlet runner, and to that circumstance we may safely ascribe its sterility. An analogous case has been long known. The vanilla plant (Vanilla planifolia) has been introduced from tropical America into India, but though it grows well, and flowers, it never fruits without artificial aid. It is the same in the hothouses of Europe. Dr. Morren, of Liége, has shown that, if artificially fertilised, every flower will produce fruit; and ascribes its sterility to the absence, in Europe and India, of some insect that in America carries the pollen from one flower to another.† When those interested in the acclimature of the natural

* "Gardener's Chronicle," Oct. 24, 1857, and Nov. 14, 1858; also T. H. Farrer, in "Annals of Natural History," Oct. 1868.

Taylor's "Annals of Natural History," vol. iii. p. 1.

productions of one country on the soil of some distant one, study the mutual relations of plants and animals, they will find that in the case of many plants it is important that the insects specially adapted for the fertilisation of their flowers should be introduced with them. Thus, if the insect or bird that assists in the fertilisation of the vanilla could be introduced into and would live in India, the growers of that plant would be relieved of much trouble, and it might be thoroughly naturalised. Judging from my experience, it would be useless to attempt the acclimature of the scarlet-runner bean in Chontales unless the humble-bee were also introduced.

Caterpillars, plant-lice, bugs, and insect pests of all kinds were numerous, and did much harm to my garden; but the greatest plague of all were the leaf-cutting ants, and I had to wage a continual warfare against them. During this contest I gained much information regarding their habits, and was successful in checking their ravages, and I shall occupy the remainder of this chapter with an account of them.

LEAF-CUTTING ANTS.-Nearly all travellers in tropical America have described the ravages of the leaf-cutting ants (Ecodoma); their crowded, well-worn paths through the forests, their ceaseless pertinacity in the spoliation of the trees-more particularly of introduced specieswhich are stripped bare and ragged, with the midribs and a few jagged points of the leaves only left. Many a young plantation of orange, mango, and lemon trees has been destroyed by them. Again and again have I been told in Nicaragua, when inquiring why no fruit-trees were grown at particular places, "It is no use planting

them; the ants eat them up." The first acquaintance a stranger generally makes with them is on encountering their paths on the outskirts of the forest crowded with the ants; one lot carrying off the pieces of leaves, each piece about the size of a sixpence, and held up vertically between the jaws of the ant; another lot hurrying along in an opposite direction empty-handed, but eager to get loaded with their leafy burdens. If he follows this last division, it will lead him to some young trees or shrubs, up which the ants mount; and then each one, stationing itself on the edge of a leaf, commences to make a circular cut, with its scissor-like jaws, from the edge, its hinder feet being the centre on which it turns. When the piece is nearly cut off, it is still stationed upon it, and it looks as though it would fall to the ground with it; but, on being finally detached, the ant is generally found to have hold of the leaf with one foot, and soon righting itself, and arranging its burden to its satisfaction, it sets off at once on its return. Following it again, it is seen to join a throng of others, each laden like itself, and, without a moment's delay, it hurries along the well-worn path. As it proceeds, other paths, each thronged with busy workers, come in from the sides, until the main road often gets to be seven or eight inches broad, and more thronged than the streets of the city of London.

It

After travelling for some hundreds of yards, often for more than half a mile, the formicarium is reached. consists of low, wide mounds of brown, clayey-looking earth, above and immediately around which the bushes have been killed by their buds and leaves having been persistently bitten off as they attempted to grow after their first defoliation. Under high trees in the thick

forest the ants do not make their nests, because, I believe, the ventilation of their underground galleries, about which they are very particular, would be interfered with, and perhaps to avoid the drip from the trees. It is on the outskirts of the forest, or around clearings, or near wide roads that let in the sun, that these formicariums are generally found. Numerous round tunnels, varying from half an inch to seven or eight inches in diameter, lead down through the mounds of earth; and many more, from some distance around, also lead underneath them. At some of the holes on the mounds ants will be seen busily at work, bringing up little pellets of earth from below, and casting them down on the everincreasing mound, so that its surface is nearly always fresh and new-looking.

Standing near the mounds, one sees from every point of the compass ant-paths leading to them, all thronged with the busy workers carrying their leafy burdens. As far as the eye can distinguish their tiny forms, troops upon troops of leaves are moving up towards the central point, and disappearing down the numerous tunnelled passages. The outgoing, empty handed hosts are partly concealed amongst the bulky burdens of the incomers, and can only be distinguished by looking closely amongst them. The ceaseless, toiling hosts impress one with their power, and one asks-What forests can stand before such invaders ? How is it that vegetation is not eaten off the face of the earth? Surely nowhere but in the tropics, where the recuperative powers of nature are immense and ever active, could such devastation be withstood.

Further acquaintance with the subject will teach the

inquirer that, just as many insects are preserved by being distasteful to insectivorous birds, so very many of the forest trees are protected from the ravages of the ants by their leaves either being distasteful to them, or unfitted for the purpose for which they are required, whilst some have special means of defence against their attacks. None of the indigenous trees appear so suitable for them as the introduced ones. Through long ages the trees and the ants of tropical America have been modified together. Varieties of plants that arose unsuitable for the ants have had an immense advantage over others that were more suitable; and thus through time every indigenous tree that has survived in the great struggle has done so because it has had originally, or has acquired, some protection against the great destroyer. The leaf-cutting ants are confined to tropical America; and we can easily understand that trees and vegetables introduced from foreign lands where these ants are unknown could not have acquired, excepting accidentally, and without any reference to the ants, any protection against their attacks, and now they are most eagerly sought by them. Amongst introduced trees, some species of even the same genus are more acceptable than others. Thus, in the orange tribe, the lime (Citrus lemonum) is less liked than the other species; it is the only one that I ever found growing really wild in Central America: and I have sometimes thought that even in the short time since the lime was first introduced, about three hundred years ago, a wild variety may have arisen, less subject to the attacks of the ants than the cultivated variety; for in many parts I saw them growing wild, and apparently not touched. The