As already mentioned, there are three principal modes in which self-fertilisation is prevented. Firstly, by the stamens and pistil being situated in different flowers, either on the same plant, or, more commonly, in different plants. These differences form the characteristics of the classes, Moncecia, Dicecia, and Poly-gamia, of Linnaeus; but it is obvious that such classes are not natural, since we have in very nearly allied species, even within the limits of what is generally considered a single genus, cases in which the one is diclinous, that is to say, has the stamens and pistil in separate flowers, while in the other, the flowers contain both.

Secondly, in other cases, the self-fertilization of plants, as was first observed by Sprengel in Epilobium angustifolium in the year 1790, is guarded against by the fact that the stamens and pistils do not ripen at the same time.

In some few cases the pistil ripens before the stamens; these species are called "proterogynous." Thus the Aristolochia has a flower which consists of a long tube with a narrow opening closed by stiff hairs which point backwards, so that it much resembles an ordinary eel-trap. Small flies enter the tube in search of honey, which from the direction of the hairs they can do easily, though on the other hand, from the same cause, it is impossible for them to return. Thus they are imprisoned in the flower; gradually, however, the pistil passes maturity, and the stigma ceases to be capable of fertilisation, while the stamens ripen and shed their pollen, by which the flies get thoroughly dusted. Then the hairs of the tube shrivel up and release the prisoners, which carry the pollen to another flower.

Again, in the common Arum, we find a somewhat similar mode of fertilisation. The well-known green leaf, as shown in the annexed diagrammatic figure (Fig. 29), encloses a central pillar which supports a number of stigmas (Fig. 29, st.) near the base, and of anthers (a) somewhat higher. Now in this case nothing would at first sight seem easier or more natural than that the pollen from the anthers should fall on, and fertilise, the pistils. This, however, is not what occurs. The stigmas mature before the anthers, and by the time the pollen is shed, have become incapable of fertilisation. It is impossible, therefore, that the plant should fertilise itself. Nor can the pollen be carried by wind. When it is shed it drops to the bottom of the tube, where it is so effectually sheltered that nothing short of a hurricane could dislodge it; and although Arum is common enough, still the chances against any of the pollen so dislodged being blown into the tube of another plant would be immense.

Diagrammatic section of Arum, h, hairs; a, anthers; st, stigmas.

Fig. 29. - Diagrammatic section of Arum, h, hairs; a, anthers; st, stigmas.

As, however, in Aristolochia, so also in Arum, small insects which, attracted by the showy central spadix, the prospect of shelter or of honey, enter the tube while the stigmas are mature, find themselves imprisoned, by the fringe of hairs (Fig. 29, h), which, while permitting their entrance, prevent them from returning. After a while, however, the period of maturity of the stigmas is over, and each secretes a drop of honey thus repaying the insects for their captivity. The anthers then ripen and shed their pollen, which falls on and adheres to the insects. Then the hairs gradually shrivel up and set the insects free, which carry the pollen with them, so that those which then visit another plant can hardly fail to deposit some of it on the stigmas. Sometimes more than a hundred small flies will be found in a single Arum. In these two cases there is obviously a great advantage in the fact that the stigmas arrive at maturity before the anthers.

Our common Scrophularia nodosa, some species of Plantago, etc, are also proterogynous, but such cases are comparatively rare.

The advantage to Scrophularia of being proterogynous, as Mr. Wilson ("Nature," September 5, 1878) has ingeniously pointed out, arises from the fact of its being fertilised by wasps, which generally begin with the upper flower and work downwards, while bees begin below and work upwards. The lower flowers are the older. Hence a bee coming from another plant of the same species fertilises the lower flowers, and then carries off a fresh supply of pollen from the upper and younger ones. On the other hand, as wasps commence from above it is an advantage that the flowers should be proter-ogynous, because the consequence is that the wasp fertilises the upper flowers, and then carries off a fresh supply of pollen from the lower and older ones.

On the other hand those in which the anthers come to maturity before the pistil are much more numerous. To the category of these plants, which are called proterandrous, belong some species of Thyme, Pinks, Epilobium (Figs. 47, 48), Geranium (Fig. 40), Malva (Figs. 43, 44), (Mallow), Impatiens, Gentians, many of the Labiatae, the Umbellifers, most of the Composites, of the Lobeliaceae, and Campanulaceae. In fact, the greater number of flowers which contain both stamens and pistil, are more or less pro-terandrous.

Fig. 30 represents a flower of the Pink in the first, or male condition. The stamens are mature, and project above the disk of the flower, while the pistil is still concealed within the tube. On the other hand Fig. 31 represents the same flower in a more advanced condition; the stamens have shrivelled up, while the pistil now occupies their place.

Again, Fig. 32 represents a flower of the Thyme (Thymus serpyllum) and shows the four mature stamens, act, and the short, as yet undeveloped pistil, p. Fig. 33, on the contrary, represents a somewhat older flower, in which the stamens are past maturity, while the pistil, p, on the other hand, is considerably elongated, and is ready for the reception of the pollen.