shapes of the primary wing-feathers, in the relative length of the hind toe, or in habits of life, as in roosting and building on trees. But the above' objection shows how completely the principle of selection has been misunderstood.

It is not

likely that characters selected by the caprice of man should resemble differences preserved under natural conditions, either from being of direct service to each species, or from standing in correlation with other modified and serviceable structures. Until man selects birds differing in the relative length of the wing-feathers or toes, etc., no sensible change in these parts should be expected.. With respect to the domestic races not roosting or building in trees, it is obvious that fanciers would never attend to or select such changes in habits."

Studies of Cultivated and Wild Plants

Still more remarkable, perhaps, is the collection of facts afforded by plants, which can be so much more easily culti vated and experimented upon than animals, while the general phenomena they present are strikingly accordant in the two kingdoms. As an example of the great mass of facts afforded by horticulture, he records that three hundred distinct varieties were produced, in the course of fifty years, from a single wild rose (Rosa spinosissima). We find in these volumes enormous collections of facts on bud-variation, or the occurrence of changes in the flower or leaf-buds of full-grown plants, from which new varieties can be and often are produced; and, after a most full and interesting discussion of the cases, it is shown that some are probably due to reversion to an ancestral form, others to reversion to one parent when the plant has been derived from a cross, and others, again, to that spontaneous variability which seems to be the universal characteristic of all living organisms.

Three very interesting chapters are then devoted to the subject of inheritance, and a host of strange and heretofore inexplicable facts are brought together, compared, and classified, and shown to be in accordance with a few general principles. Then follow five chapters on crossing and hybridism, perhaps the most important in the whole work, since they afford the clue to so much of the varied structure and complex relations of animals and plants. Notwithstanding the

enormous mass of facts and observations here given, the portion relating to plants is often but an abstract of the results of his own elaborate experiments, carried on for a long series of years, and given at length in three separate volumes on The Fertilisation of Orchids, on Cross and Self-Fertilisation of Plants, and on The Forms of Flowers. These works may be said to have revolutionised the science of botany, since, for the first time, they gave a clear and intelligible reason for the existence of that wonderful diversity in the form, colours, and structure of flowers, on the details of which the systematic botanist had founded his generic and specific distinctions, but as to whose meaning or use he was, for the most part, profoundly ignorant. The investigation of the whole subject of crossing and hybridity had shown that, although hybrids between distinct species usually produced sterile offspring, yet crosses between slightly different varieties led to increased fertility; and, during some experiments on this subject, Darwin found that the produce of these crosses were also remarkable for vigour of growth. This led to a long series of experimental researches, the general result of which was to establish the important proposition that cross-fertilisation is of the greatest importance to the health, vigour, and fertility of plants. The fact that the majority of flowers are hermaphrodite, and appear to be adapted for self-fertilisation, seemed to be opposed to this view, till it was found that, in almost every case, there were special arrangements for ensuring, either constantly or occasionally, the transference of pollen from the flowers of one plant to those of another of the same species. In the case of orchids, it was shown that those strange and beautiful flowers owed their singular and often fantastic forms and exceptional structure to special adaptations for cross-fertilisation by insects, without the agency of which most of them would be absolutely sterile. Many of the species are so minutely adapted to particular species or groups of insects, that they can be fertilised by no others; and careful experiment and much thought was often required to find out the exact mode in which this was effected. some instances the structure of the flowers seemed adapted to prevent fertilisation altogether, till it was at length discovered that a particular insect entering the flower in one particular

In

way caused the pollen to stick to some part of its body, which was always the exact part which the insect, on visiting another flower, would bring in contact with the stigma, and thus fertilise it. These investigations explained a host of curious facts which had hitherto been facts only without meaning, such as the twisting of the ovary in most of our wild orchids, which was found to be often necessary to bring the flower into a proper position for fertilisation,—the existence of sacs, cups, or spurs, the latter often of enormous length, but shown to be each adapted to the structure of some particular insect, and often serving to prevent other insects from reaching the nectar which they might rob without fertilising the flower, the form, size, position, rugosities, or colour of the lip, serving as a landing-place for insects and a guide to the nectar-secreting organs,-the varied odours, sometimes emitted by day, and sometimes by night only, according as the fertilising insect was diurnal or nocturnal, and other characters too numerous to refer to here, so that it. became evident that every peculiarity of these wonderful plants, in form or structure, in colour or marking, in the smoothness, rugosity, or hairiness of parts of the flower, in their times of opening, their movements, or their odours, had every one of them a purpose, and were, in some way or other, adapted to secure the fertilisation of the flower and the preservation of the species.

Researches on the Cowslip, Primrose, and Loosestrife

The next set of observations, on some of our commonest English flowers of apparently simple structure, were not less original and instructive. The cowslip (Primula veris) has two kinds of flowers in nearly equal proportions: in the one the stamens are long and the style short, and in the other the reverse, so that in the one the stamens are visible at the mouth of the tube of the flower, in the other the stigma occupies the same place, while the stamens are half-way down the tube. This fact had been known to botanists for seventy years, but had been classed as a case of mere variability, and therefore considered to be of no importance. In 1860 Darwin set to work to find out what it meant, since, according to his views, a definite variation like this must have a purpose.

After a considerable amount of observation and experiment, he found that bees and moths visited the flowers, and that their probosces became covered with pollen while sucking up the nectar, and further, that the pollen of a long-stamened plant would be most surely deposited on the stigma of the long-styled plants, and vice versa. Now followed a long series of experiments, in which cowslips were fertilised either with pollen from the same kind or from a different kind of flower, and the invariable result was that the crosses between the two different kinds of flowers produced more good capsules, and more seeds in each capsule; and as these crosses would be most frequently effected by insects, it was clear that this curious arrangement directly served to increase the fertility of this common plant.

The same thing was found to occur in the primrose, and in many other species of primulaceæ, as well as in flax (Linum perenne), lungworts (Pulmonaria), and a host of other plants, including the American partridge-berry (Mitchella repens). These are called dimorphic heterostyled plants.

Still more extraordinary is the case of the common loosestrife (Lythrum salicaria), which has both stamens and styles of three distinct lengths, each flower having two sets of stamens and one style, all of different lengths, and arranged in three different ways: (1) a short style, with six medium and six long stamens; (2) a medium style, with six short and six long stamens; (3) a long style, with six medium and six short stamens. These flowers can be fertilised in eighteen distinct ways, necessitating a vast number of experiments, the result being, as in the case of the cowslip, that flowers fertilised by the pollen from stamens of the same length as the styles, gave on the average a larger number of capsules and a very much larger number of seeds than in any other The exact correspondence in the length of the style of each form with that of the stamens in the two other forms ensures that the pollen attached to any part of the body of an insect shall be applied to a style of the same length on another plant, and there is thus a triple chance of the maximum of fertility. Some other species of lythrum, of oxalis, and pontederia, were also found to have three-formed stamens and styles; and in the case of the oxalis, experiments were

case.

made showing that crosses between flowers with stamens and styles of unequal length were always nearly barren. During these experiments 20,000 seeds of Lythrum salicaria were counted under the microscope. For several years a further supplementary series of experiments were carried out, showing that the seeds produced by the illegitimate crosses (as he terms them) were not only very few, but, when sown, always produced comparatively weak, small, or unhealthy plants, not likely to exist in competition with the stronger offspring of legitimate crosses. There is thus the clearest proof that these complex arrangements have the important end of securing both a more abundant and more vigorous offspring.

Perhaps no researches in the whole course of the study of nature have been so fertile in results as these. No sooner were they made known than observers set to work in every part of the world to examine familiar plants under this new aspect. With very few exceptions it is now found that every flower presents arrangements for securing cross-fertilisation, either constantly or occasionally, sometimes by the agency of the wind, but more frequently through the mediation of insects or birds. Almost all the irregularity and want of symmetry in the forms of flowers, which add so much to their variety and beauty, are found to be due to this cause; the production of nectar and the various nectar-secreting organs is directly due to it, as are the various odours and the various colours and markings of flowers. In many cases flowers which seem so simply constructed that the pollen must fall on the stigma and thus produce self-fertilisation, are yet surely crossfertilised, owing to the circumstance of the stigma and the anthers arriving at maturity at slightly different periods, so that, though the pollen may fall on the stigma of its own flower, fertilisation does not result; but when insects carry the pollen to another plant the flowers of which are a little more advanced, cross-fertilisation is effected. There is literally no end to the subjects of inquiry thus opened up, since every single species, and even many varieties of flowering plants, present slight peculiarities which modify to some extent their mode of fertilisation. This is well shown by the remarkable observations of the German botanist Kerner, who points out that a vast number of details in the structure of