duction. Sooner or later, the anthers of the stamens open in one way or another so as to allow the pollen to escape; and, viewed under the microscope, the pollen-grains are seen to vary greatly in size and form in different species of plants. The grains of pollen may be round (Fig. 220) or oval in form; in the evening primrose (Fig. 222) and fuchsia, they are of triangular shape; in the hollyhock and melon (Fig. 223) they are spinous; and in the orchids they are united to form masses (Fig. 221) called pollinia.

The pollen-grains being conveyed to the stigma (Fig. 218, sg) of the pistil, they are there attached by the aid of a glutinous secretion, which

may likewise be credited with a specific influence on the pollengrains, in that it appears to stimulate the curious development they next evince. This development consists in the rupture of the outer of the two layers of which each pollen-grain consists. Through the ruptured outer coat, the inner layer begins to grow in the form of a long tube-the pollen-tube (Fig. 220)-which penetrates the tissue of the style (Fig. 224), and grows downwards to reach the ovules contained in the ovary. In some plants, the pollen-tubes emitted from

one pollen-grain may be very numerous, although as a rule only one tube grows from each grain.

Now, the essence of fertilisation (ie. the production of a "seed" fitted to produce a new plant) appears to consist in the contact of the pollen-tube with the ovule, so that the viscid matter called fovilla, contained within the pollen-grain, may be applied to the structures of the ovule. The most important part of the ovule itself is a small cellular body called the nucleus, enveloped in a couple of coverings. The hollow interior of the nucleus is named the embryo-sac, and an opening

FIG. 224.-POLLEN-TUBES OF DATURA PENETRATING THE STYLE (MAGNIFIED).

called the micropyle also exists in the coats of the ovule. Through this opening the pollen-tube passes, gaining admittance thereby to the nucleus, and thence to its hollow body or embryo-sac, wherein the fovilla, or contents of the pollen-grain, are discharged.

Such is the work of fertilisation, and such are the processes in virtue of which the ovule becomes the seed. As the result of these processes,

the "embryo," or young plant, is duly formed within the embryo-sac, and thus, even before the seed is planted, development has already proceeded to a certain extent. In the seed of a pea or bean (Fig. 225), for instance, we readily perceive the rudiment of the stem (p), the beginning of the root (r), and likewise the first appendages or "seed leaves (c)," which that stem will develop.

p r

The process of fertilisation, thus described in its essential nature, involves in the case of certain plants some curious details, the mere mention of which may stimulate to an independent research into botanical lore. Thus, often the pollen-tubes may require, from the length of the style of the pistil, to grow to a large relative extent. In the crocus, the pollen-tube requires to grow to a length of three inches before it can reach the ovules FIG. 225.-SECTION OF BEAN. in the ovary. The number of pollen-grains in flowers may be apparently in excess of all reasonable proportions-a fact to be accounted for on the wellfounded idea that the pollen of a flower is not usually limited to that particular flower's wants, but may be destined to serve for the fertilisation of others of the same species. In the great flowered cactus (Cactus grandiflorus), Morren says there are about 500 anthers, 24 stigmas, and 30,000 ovules. Assuming that each anther contains 500 pollen-grains, this will give a total of 250,000 grains to each flower; and the interval or space between the stigma and the ovules of this plant is about 1,150 times the diameter of the pollen-grains. Nature appears exceedingly lavish in her development of pollen. If the Tennysonian aphorism that

be true as it unquestionably is-the apparent over-production of pollen-grains is even more remarkable, although we have to take into account the fact just noted, that the development of pollen bears a relation rather to the species and race, than to the individual necessities of the plant. Otherwise, Fritz Müller's estimate, that in a single flower of Maxillaria there are developed 34,000,000 grains of pollen, must present itself as an inexplicable fact of botanical science. Even the wheat-plant produces about 50 lbs. of pollen to the acre. The pollen of the cone-bearing plants (Conifera), such as the firs, larches, pines, or that of the catkin-bearers (Amentifera), is often borne through the air as showers of yellow sulphur-like dust. This dust, falling in regions where the elements of botany are unknown, cause perturbation amongst the unlearned, and result in the penning of

epistles to "Mr. Editor" by way of inquiry whether or not the sulphureous shower is a portent or grave omen of coming disaster or impending peril.

The phenomena of fertilisation just detailed, take place in our primrose, as in all ordinary plants; but whilst there exists a uniformity in the details of this process, there is also found a literally amazing variety in the fashions whereby pollen is conveyed to the stigma of the pistil. Once placed in the natural position for fertilisation, the growth of the pollen-tube follows as a matter of course. But the means whereby the pollen reaches the stigma, and the various fashions in which it may gain its ultimate position on the pistil, constitute features in which are bound up some of the most important issues of plant existence. To rightly comprehend the bearing of fertilisation, a glance at our wallflower (Fig. 205), primrose (Fig. 208), foxglove (Fig. 206), or buttercup will suffice as a starting-point for further investigation. Within the primrose and the buttercup are situated, as we have seen, the two sets of organs-stamens and pistil-necessary to secure the production of seed and the continuance of the race. Hence it might form a very natural and reasonable inference, that the pollen from the numerous stamens of a buttercup flower should be used to fertilise the ovules of the pistil of that flower. Such a process-that in which a flower's own pollen is used to fertilise its own ovules-is termed "self-fertilisation." Looking at the vast majority of our flowers and plants, which possess each a perfect array of stamens and pistil, the normal course of things seems strongly suggestive of self-fertilisation. Hence, in the early days of botany, self-fertilisation was undoubtedly believed to be the rule of nature. Now, there can be no question whatever that "self-fertilisation " does occur in nature, but there is as little doubt that it is the exception, and not-as botanists from the days of Linnæus well-nigh to our own day have maintained the rule, of plant life. There can be little doubt, for instance, that many small species of the buttercup order (Ranunculacea-e.g. Ranunculus hederaceus) are self-fertilised, because we find the stamens to arch over the pistil, and to shed their pollen on the carpels. In Agrimonia, in the same order, the stamens, at first curved outwards, curve inwards, so as to bring the pollen within easy reach of the stigmas. So, also, in a species of Malvaceæ (Malva rotundifolia), Müller has demonstrated that this plant is self-fertilised, since stigmas and anthers actually intertwine, and are thus placed in the most favourable position for the fertilisation of the ovules. Some species of Geraniacea (e.g. Geranium pusillum) are self-fertilising likewise; and many flowers belonging to the rose tribe (Rosacea), such as Potentilla, fertilise themselves.

It is a remarkable fact that in certain plants (eg. many violets; Lamium amplexicaule; Oxalis, &c.) very small, inconspicuous, and

closed flowers are produced, in addition to the ordinary conspicuous and, as we shall see, "cross" or insect-fertilised flowers. These closed flowers have been named "cleistogamous "-a term applied by Kuhn in 1867. They are self-fertilised, and produce numerous seeds; and their occurrence in the same plant along with crossfertilised blossoms, may perhaps be best explained on the theory that, whilst the ordinary and less fertile flowers will afford to the plant the advantages and benefits which accrue from "cross-fertilisation," the "cleistogamous" flowers may be regarded as the normal means for the ordinary increase of the race. What the flower loses in variation by the sparing fertility of the cross-fertilised flowers, it may gain in the number of seeds which the cleistogamous flowers produce. Cleistogamous flowers likewise tend to economise pollen. Whilst 400 pollen-grains may serve the purpose of close or selffertilisation in Oxalis, or even 100 grains in some violets, three-anda-half million grains may be produced in the insect-fertilised flowers of the peony, and many millions in the case of wind-fertilised flowers, whose pollen, like that of the firs, has to be distributed over immense areas of land.

There appears, therefore, to be a proportion of plants in which the existence of stamens and pistil in the same flower-the normal condition of matters in ordinary plants-is meant to and does secure the fertilisation of the ovules by the flower's own pollen. Why, then, seeing that the presence of correlated stamens and pistils in each flower appears to be a common condition of plant life, do we assume that not self-fertilisation but the opposite process-crossfertilisation is the rule of nature? The reply to this query involves more than one important consideration. Let us briefly endeavour to find a convenient starting-point in the familiar flower which Peter Bell despised, and which, to minds of utilitarian type amongst ourselves, is but a primrose still, and "nothing more."

If we study the structure of the primroses we may gather in a bed of these flowers, it will be found that the blossoms obtained from one set of plants will vary in certain respects from the flowers of the other and neighbouring plants. There is no difference in appearance or in outward aspect between the primroses, because the differences referred to affect chiefly the position of the stamens and the length of the style (or "neck" of the pistil) in each variety. But we may readily discover that, selecting any one primrose plant, all the flowers of that plant will be either long-styled (Fig. 208 A) or short, styled (Fig. 208 B), and will not exhibit a mixture of the two varieties. "The two kinds of flowers," says Mr. Darwin, speaking of the long and short-styled cowslips, which form a closely allied species to the primroses, are never found on the same individual plant;" and he also remarks that he has never met with any