If it be objected that I am assuming the existence of these gradual modifications, I must reply that it is not here my purpose to discuss the doctrine of Natural Selection. I may, however, remind the reader that Mr. Darwin's theory is based on the following considerations: - I. That no two animals or plants in nature are identical in all respects. 2. That the offspring tend to inherit the peculiarities of their parents. 3. That of those which come into existence, only a certain number reach maturity. 4. That those which are, on the whole, best adapted to the circumstances in which they are placed, are most likely to leave descendants.

Now, applying these considerations to flowers; if it be an advantage to them that they should be visited by insects (and that this is so will presently be shown), then it is obvious that those flowers which, either by their larger size, or brighter colour, or sweeter scent, or greater richness in honey, are most attractive to insects, will, caeteris paribus, have an advantage in the struggle for existence, and be most likely to perpetuate their race.

Every garden indeed is a sufficient proof that in size and colour, flowers are susceptible of great modifications; and insects unconsciously produce changes similar to those which man effects by design.

Insects are useful to plants in various ways. Thus, a species of acacia mentioned by Mr. Belt,1 if unprotected, is apt to be stripped of its leaves by a leaf-cutting ant, which uses the leaves, not directly for food, but, according to Mr. Belt, to grow mushrooms on. The acacia, however, bears hollow thorns, and each leaflet produces honey in a crater-formed gland at the base, and a small, sweet, pear-shaped body at the tip. In consequence, it is inhabited by myriads of a small ant, Pseudomyrma bicolor, which nests in the hollow thorns, and thus finds meat, drink, and lodging all provided for it. These ants are continually roaming over the plant, and constitute a most efficient body-guard, not only driving off the leaf-cutting ants, but even in Mr. Belt's opinion, rendering the leaves less liable to be eaten by herbivorous mammalia.

The principal service, however, which insects perform for plants is that of transferring the pollen from one flower to another.

I will not now enter on the large question why this cross-fertilisation should be an advantage; but that it is so has been clearly proved. Kolreuter speaks with astonishment of the "statura portentosa" of some plants thus raised by him; indeed, says Mr. Darwin ("Animals and Plants under Domestication," ch. xvii.), "all experimenters have been struck with the wonderful vigour, height, size, tenacity of life, precocity, and hardiness of their hybrid productions." Mr. Darwin himself, however, was, I believe, the first to show that if a flower be fertilised by pollen from a different plant, the seedlings so produced are much stronger than if the plant be fertilised by its own pollen. I have had the advantage of seeing several of these experiments, and the difference is certainly most striking. For instance, six crossed and six self-fertilised seeds of Ipomcea purpurea were grown in pairs on opposite sides of the same pots; the former reached a height of 7 ft., while the others were on an average only 5 ft. 4 in. The first also flowered more profusely. It is moreover remarkable that in many cases plants are themselves more fertile if supplied with pollen from a different flower, a different variety, or even, as it would appear in some instances (in the passion flower, for instance), from a different species. Nay, in some cases pollen has no effect whatever unless transferred to a different flower. Fritz. Muller has recorded some species in which oollen, if placed on the stigma of the same flower, has not only no more effect than so much inorganic dust; but, which is perhaps even more extraordinary, in others, he states that the pollen placed on the stigma of its own flower acted on it like a poison. This he noticed in several species: the flower faded and fell off; the pollen-grains themselves, and the stigma in contact with them, shrivelled up, turned brown, and decayed; while other flowers on the same branch, which were not so treated, retained their freshness.

1 F. Muller has observed similar facts in Sta. Catharina. (Nature, vol. x. p. 102.)

The transference of the pollen from one flower to another is, as I have already mentioned, effected principally either by the wind or by insects. In the former case the flower is rarely conspicuous; indeed Mr. Darwin finds it "an invariable rule that when a flower is fertilised by the wind it never has a gaily-coloured corolla." Conifers, grasses, birches, poplars, etc, belong to this category.

In such plants a much larger quantity of pollen is required than where fertilisation is effected by insects. Everyone has observed the showers of yellow pollen produced by the Scotch fir. It is an advantage to these plants to flower before the leaves are out, because the latter would greatly interfere with the access of the pollen to the female flower. Hence such plants, as a rule, flower early in the spring. Again, in such flowers the filaments of the stamens are generally long, and the pollen is less adherent, so that it can easily be detached by the wind, which would manifestly be a disadvantage in the case of those flowers which are fertilised by insects. On the other hand, it is an advantage to most seeds to be somewhat tightly attached, because they are then only removed by a high wind which is capable of carrying them some distance. I say "to most" because this does not apply to such seeds as those of the dandelion, which are specially adapted to be carried by the wind.

Again, as Mr. Darwin has pointed out, irregular flowers appear to be almost always fertilised by insects.

Wind-fertilised flowers, moreover, generally have the stigma more or less branched or hairy, which evidently tends to increase its chance of catching the pollen.