This is a large family, consisting of fourteen genera, and contains: Veronica (Fig. 97), Verbascum (Mullein), (Fig. 98), Linaria, Antirrhinum (Snapdragon), Scro-phularia (Fig. 99), Digitalis (Foxglove), (Fig. 100), Euphrasia (Eyebright), (Fig. 106), Rhinanthus (Rattle), etc.

The first two genera have more or less open flowers; while the others are more distinctly tubular, and have much the appearance of Labiatae, but differ from that group in having the ovary two-celled, with several ovules in each cell.

Veronica. The flowers are rendered conspicuous by their colour and the association in racemes. In V. Chamcedrys (Fig. 97), the anthers and stigmas ripen simultaneously, but while the latter project straight forwards, the two stamens turn outwards, so that fertilisation can hardly take place.

Veronica Chamcedrys.

Fig. 97. - Veronica Chamcedrys.

V. Beccabunga in many respects resembles V. Chamcedrys; but is proterogynous. In V. spicata some flowers are proterogynous, others proterandrous, and being, in consequence of their conspicuousness, much visited by insects, they appear to have lost the power of self-fertilisation. In V. hedercefolia, on the contrary, the flowers are minute, and habitually fertilise themselves.

The species of Verbascum (Mullein) are showy plants, with either white or yellow flowers, forming a tall spike, which in V. Thapsus reaches a height of four feet. V. nigrum, L. (Fig. 98) has yellow flowers; the stamens clothed with beautiful violet hairs. They secrete very little honey, but are visited by various insects for the sake of the pollen, and perhaps also on account of the glandular terminations of the violet staminal hairs. The stamens turn somewhat upwards, the pistil, on the contrary, downwards, so that an insect alighting on the lower lip of the corolla, which is the most convenient place, would naturally come in contact with it before touching; the stamens. V. nigrum, however, according to Gaertner, cannot be fertilised by its own pollen.

Verbasauu Tliapsus.

Fig. 98. - Verbasauu Tliapsus.

The genus Scrophularia, from which the family takes its name, is remarkable in many respects. From the general arrangement of the blossom in flowers of the Labiate form, the pistil could hardly occupy any other position than the central median line, and a fifth stamen would accordingly be in the way. It has therefore disappeared, though Muller mentions that he once found one in Laminm album. In Scr. nodosa (Fig. 99), however, the four normal stamens and the pistil occupy the lower side of the flower, and the presence of a fifth stamen, even if useless, is under these circumstances not injurious. A rudimentary fifth stamen is, in fact, habitually present, and in some cases bears pollen. Scr. nodosa is proterogynous, and is much frequented and fertilised by wasps. Pentstemon also has a fifth stamen, which curves in a very curious manner from the upper to the under side of the flower so as to be out of the way of the pistil. Ogle regards it as perfectly useless (Popular Science Reviezu, Jan. 1870), but it is so large that I cannot help thinking it must have some function, though I am unable to suggest one.

Fig. 100. - Section of Digitalis purpurea, showing the anthers unripe and horizontal. Fig. 101. - Ditto, more advanced. The upper anthers ripe and vertical, the lower ones as before. Fig. 102. - Ditto, still more advanced. All the anthers ripe and vertical.

Scrophularia nodosa.

Fig. 99. - Scrophularia nodosa.

Scrophulariaceae 109Scrophulariaceae 110

Fig. 101

Scrophulariaceae 111

Fig. 102.

In Linaria vulgaris the flowers form a closed box terminating behind in a spur, 10 - 13 mm. in length, which contains the honey, and the orifice of which is protected by hairs. Under these circumstances, the long-lipped bees are the only insects which can suck the honey. Antirrhinum majus (the Snapdragon) differs in the larger size of the flowers, the greater firmness with which they are closed, and in the position of the honey, which lies at the basis of the corolla, and does not penetrate into the short spur, which is hairy, and therefore not suited for such a purpose. They are almost always fertilised by humble bees, though smaller bees occasionally force their way into them.

Digitalis purpurea (the Foxglove) is also exclusively fertilised by humble bees, which alone are large enough to fill the bell, and thus to deposit pollen on the stigma. The flower is proterandrous, but appears to be self-fertile if the visits of humble bees are delayed or prevented. The cells of the anthers, as Ogle has pointed out, are at first transverse (Fig. 100), but as the two pairs ripen they successively become longitudinal (Figs. 101 and 102).

Bartsia odontites.

Fig. 103. - Bartsia odontites.

The other British genera of this group have narrow tubular flowers; in which the upper lip protects the anthers and pistil, while the lower lip serves as an alighting stage for insects. The stamens are so arranged that the insects in searching for the honey dust themselves with the pollen. For instance, in Bartsia odontites (Fig. 103), the common red Bartsia, the flower forms a tube 4 - 5 mm. long; at the base of which is the honey, while the entrance is protected against rain by the four hairy anthers. These lie close together; but immediately below them, the filaments of the stamens separate so as to leave a space through which bees can insert their proboscis, and thus reach the honey. In doing so they naturally dust themselves with pollen, some of which they transfer to the stigma (Fig. 105, st) of the next flower they may visit. Muler has observed that in plants of this species which live in shady places and are consequently less visited by insects, the pistil is shorter (Fig. 104), the stigma consequently nearer to the anthers, and more likely to be fertilised directly by them.

Bartsia odontites. Flower with a short pistil.

Fig. 104. - Bartsia odontites. Flower with a short pistil.

Ditto. Flower with a long pistil.

Fig. 105. Ditto. Flower with a long pistil.

He also observes that this flower is not perfectly adapted to present circumstances, since bees are able to, and often do, insert their proboscis above the stamens, in which case they do not fertilise the flower.

Euphrasia officinalis (the Eyebright) (Fig-. 106), agrees in many respects with the preceding; but there is no room above the stamens for the proboscis of the bee. The anthers (Fig. 107) also, which in Bartsia odontites are merely locked together by hair, in this species are more intimately connected, the two uppermost anthers to one another, the lower anther of each upper pair with the upper anther of the lower stamen on the same side. The lower anther of the lower stamen is produced into a strong point (Fig.

107, which is touched by the proboscis of the bee in passing down the tube to the nectary, and serves as a lever, shaking the whole system of anthers and thus causing the pollen to fall outon to the bee.

Euphrasia officinalis.

Fig. 106. - Euphrasia officinalis.

Flower of Euphrasia officinalis.

Fig. 107. - Flower of Euphrasia officinalis.

In this species also H. Muller has observed that there are two forms, a larger one which is adapted to be fertilised by insects, and a smaller one which more frequently fertilises itself.

In Rhinanthus Cristagalli (the Common Rattle) the anthers are locked together, and the pollen is shed on to the bee, but the mode in which this is effected is not the same. In this species, as in Bartsia odontites, the bee has to pass its proboscis between the filaments of the anthers in order to reach the honey, and the space between them is so narrow, that the bee in pressing its proboscis down the tube, presses the filaments apart, thus shaking the anthers, and freeing the pollen. In this species also H. Muller has observed the existence of two forms.

In the common Pedicularis (Fig. 108) (Pedicularis sylvatica), which has been well described by Hilde-brand and Delpino, the arrangement is somewhat different. The anthers open on their inner sides, and the edges of the open anther cells on the one side of the flower exactly correspond with, and are applied to, the corresponding edges of the anthers on the other side of the flower; each pair of anthers thus forming as it were, a closed box. The outer sides of the anthers are slightly attached to the walls of the hood. But the sides of the hood are somewhat too near together to admit the head of the humble-bee, and the insect therefore, in order to reach the honey, pushes them a little apart, thus opening the anther-box and letting down a little shower of pollen, which is prevented from spreading by the fringe of hairs on the lower edge of the anther, and thus falls on to the head of the bee, at the very spot which a moment before had touched the stigma, and which will again touch that of the next flower she visits.

Pedicularis sylvatica.

Fig. 108. - Pedicularis sylvatica.

In P. palnstris the point m is elongated, and the anthers, in the specimens which I have examined, are glabrous.

The structure of Melampyrum agrees in essentials with that of Pedicularis. In Calceolaria pinnata, Hildebrand describes an arrangement somewhat similar to that which we shall meet with in Salvia.