numerously beset, still show their purple hue at the spots where the cells join (fig. 3.)

Common garden Sage leaves, and many other plants of the same family, have similar glands; those of the Sage being white. In the Coleus glands there is a tendency to division into four parts, shown in fresh specimens by the cross-like depression. In a Digitalis (foxglove) from Calcutta, four-celled glands containing a white material are very conspicuous.

Let us now pass to another sort of structure, in which great variety of development may be easily seen-namely, anthers with their filaments or stalks. Everybody notices anthers like those of lilies, with their abundance of coloured pollen; but few think how the pollen is formed, and few investigate the structure of anthers that need the microscope for their elucidation. Pollen grains are formed in special cells in the substance of the anther, and they must, when ready for their work, be able to get out. This is usually accomplished by the bursting of the anther; but very interesting exceptions are found. Fig. 5 shows the anthers of an Azalea; twin bottles, with round mouths opening at the right time. From one of these bottles the pollen is in the act of escaping.

Anthers in this stage may be well preserved in cells with glycerine jelly. Each filament carries a pair of bottles, and the figure shows them in two positions-with their mouths upwards in one case, and downwards in the other. Both positions may be seen in the living plant.

Fig. 6 exhibits Rhododendron anthers. These, again, are pairs of long bottles composed of a very delicate membrane that splits lengthwise with great facility. The mouths of these bottles are less firm and regular than those of the Azalea, that are strengthened by a thick rim, and somewhat remind one of moss capsules.

In Berberis Darwinii * (fig. 7), the bottle form is abandoned for a sort of box with a lid, which opens when the pollen is ripe. In this case each anther cell splits twice lengthwise, and the lid is formed of a strip left in adhesion only at the top, and free to fall open.

In the Narrow Leaf Bay* (fig. 8), the box form again occurs, much like that of the Berberis; but the filament, or stalk of the anther, is furnished with two projecting wings, supposed to be glandular. We shall have more to say about these wings presently, but will now advise the student to examine the anthers of various plants in different stages of growth, making their sections in longitudinal and transverse directions, gently flattening them in a drop of water

* The peculiarity is not confined to this species or variety.

on a slide under a covering glass, and viewing with onequarter or higher powers. The mode of growth of the pollen buds will then be seen, and it will be found that the cells below the epidermis often present curious and beautiful appearances of spiral or other fibres. The Siberian Squill, common in gardens (Scilla Sibirica), and as a pot plant, shows the spiral fibres something like those in the leaves of Sphagnum exceedingly well. "Micrographic Dictionary gives, amongst others, Narcissus poeticus, Populus alba, and Datura Stramonium, as having spiral fibres; Iris florentina, Hyacinthus orientalis, and Convallaria, as having annular fibres; Tritillaria imperialis (internal face), and Viola odorata as having reticulated fibres; Nuphar lutea, Bryonia dioica, Primula sinensis, and Lupinus, as having arched fibres, on three sides of the cells, the fourth free. Trumpet Lilies, Calceolarias, and Larkspurs, should also be examined, as their fibres stand upright. An investigation of a dozen or two of flowers easily found in a garden or conservatory would well repay the trouble.

Botanists regard all the outgrowths from the stems of plants as leaves of some kind, however they may be modified in form or colour. What are commonly known as "leaves" of course retain that appellation in scientific works; but sepals, petals, anthers, &c., are all morphologically considered as modified leaves. At first this sort of philosophy appears exceedingly far-fetched and difficult; but there are several easy modes of getting a sufficient insight into the matter to show that it is true in certain groups of cases; and when the student has got thus far, he need not be in a hurry to doubt that with greater knowledge he would find it true in others. In regarding various parts on modified leaves the notion must be avoided that a true leaf is made first, and the particular parts made out of it afterwards by modification. This would be quite wrong. What is considered is, that the various parts begin as leaves begin, and that their growth is modified as it goes on, the cells of which they are composed arranging themselves in given directions and quantities, according to the nature of the patterns ultimately required. Nothing can appear more different than the ordinary leaf of a plant and its stamens with their filaments. But in Buttercups, Bachelor's Buttons, &c., a quantity of stamens pass into the condition of flower-leaves or petals; and if a wild single Rose is compared with a full-leaved garden variety, it will be found that the "doubling" has consisted in the development of the anthers and filaments into floral leaves. The elegant double Cherry that flowers in Spring is another illustration of stamens taking the petal form.

Minor modifications of the filament are also common, and

fig. 9 shows the number of ornamental expansions that grow down those of the Sparmannia Africana, a shrub belonging to the lime-tree family, coming from Africa, and not uncommon in greenhouses, or as an out-door plant in summer. The flower of Sparmannia owes much of its beauty to its modified filaments. It has four white petals, narrow at the base and broad at the rim, arranged in an open cross. Between these shine four white spear-shaped sepals, and the centre is occupied by an immense number of filaments, the outer golden yellow, and barren; the inner ones longer, rich red in colour and bearing purplish anthers, and both sets of filaments are provided with a series of expansions, as shown in fig. 9a, the whole effect being very pretty.

In this instance we find that the filaments, instead of being simple stalks, exhibit lateral growths; and having in this and in the case of the Berbery filaments witnessed such developments, we shall not be surprised in other plants to find much more accomplished in the way of modification.

If we look at the Cannas, now common as greenhouse plants, and planted out in summer to give sub-tropical effects, we notice at once a striking departure from ordinary flowers, and at first may not perceive in the narrow petal-like objects anything that will do for either stamen or pistil. We shall, however, on careful examination, see that some of the narrow petals have whitish tips and pollen masses upon them. These are the stigmas, as shown in fig. 10A. The anthers are of similar character, looking like leaves. Their filaments (stalks) are expanded on one side in a petal, and the actual anther, with a slight expansion on the other side, will be noticed arranged longitudinally near the top (see fig. 10).

The Cannas belong to the Marantaceæ, in which the plants yielding the true arrow-root are found, and botanists describe them as having "a calyx of three sepals, a corolla of six pieces, five of which are erect, the other reflexed; these may be considered rather as abortive stamens than petals; the one fertile stamen is petal-like, with an anther on the margin;, the style is also petal-like, with a linear stigma.' The seed of these plants is very hard, and justifies the popular name of "Indian Shot."

* 66

My acquaintance with Cannas is confined to the greenhouse varieties, known as " Alexandra," &c., and they appear to conform very closely in structure to the figures of C. Indicus and others in Roscoe's great work on the Scitamineæ. When the flower has opened, any casual observer would at first be puzzled to find the reproductive organs. He would see a mass

* Treasury of Botany.

of pollen adhering to what looks like a slender leaf, and might detect no symptoms of a stigma. He would see an anther-like body attached to the margin of another leaf, but most likely without a trace of pollen upon it. These appearances would be puzzling, but easily explained by unrolling a few buds, just before their natural time of opening. It would then be seen that the stigma was a whitish body at the apex of the leaf (or petaloid expansion), and that it was soft and full of fluid. Thinly slicing it, and placing it under a 1-inch power, would show the structure exhibited in fig. 10-a multitude of fine tubes. The anther grows so as to touch the stigma-leaf; and frequently, if not always, transfers the whole of its pollen to that body before the bud opens. This had happened in every case I have examined. Soon after the opening of the flower the stigma shrivels, and often leaves scarcely a trace to be seen. The impregnation thus appears to occur in the bud, and unless insects have any action in carrying pollen from one plant to another, the greater part of the pollen-globules of any plant have no chance of performing their function, as they are deposited below the stigma, and cannot get at it by any movement of their own. The soft tubular filaments catch and entangle the pollen that is used.

The petaloid condition of styles and of stamens, which is normal in some plants, occurs in others as a monstrosity. For example, I noticed, in February, a Fuchsia (in a greenhouse) with a very strange-looking flower. Two of the sepals, instead of taking their usual appearance, resembling white petals, had grown into green leaves, and another had formed itself into a hollow slipper-like object, while some of the stamens had developed petaloid outgrowths, as shown in fig. 11, with rich red colours.

Facts of this kind are very instructive, because they indicate modes of transition by which one group of plants might give rise to varieties that would be classed in other genera, families, or orders. There is an obvious difference between variations from the parent type that are properly called deformities, and which involve some injury or infirmity, and those that are consistent with vigorous growth and reproduction under certain conditions. If means could be found of perpetuating such an abnormal Fuchsia as I have described, we should have a new garden flower. In this Fuchsia, the outgrowth was from the anther itself, not from the filament. In Dr. Maxwell Masters' "Vegetable Teratology," a long list is given of plants subject to "petalody of the stamens;

Ray Soc. publication.

and amongst them it appears to occur most commonly in Ranunculus, Anemone, Poppy, Clematis, Hepatica, Geranium, Pelargonium, Camellia, Deutzia, Fuchsia, &c., those mentioned being within easy observation of frequenters of gardens or cultivators of flowers. Dr. Masters says: Dr. Masters says: "When petalody specially affects the anther-lobes, as in Arbutus, Petunia, Fuchsia, &c., the venation of the petal-like portion is very frequently laminar, thus tending to show that the anther is in such cases really a modification of the blade of the leaf; but as, on the other hand, we often find petal-like filaments having pollen sacs on their sides, it is clear we must not attribute the formation of pollen to the blade of the leaf only, but we must admit that it may be formed by the filament as well." He also says that the "tendency to petalification is greater among those plants which have their floral elements arranged in a spiral series, than amongst those where the verticillate arrangement exists."

Doubtless the cultivators of "Popular Science" will look out for instances of these and other modifications; and they may be assisted by another quotation from Dr. Masters, who informs us that De Candolle observes that, in the Ranunculacæ, the species of Clematis become double by the expansion of the filament, those of Ranunculus by the dilatation of the anthers, and those of Helleborus by the petal-like development of both filament and anther."

It would have been easy to have thrown the matter of the preceding pages into a more regular form, but the one chosen has arisen out of accident that may happen to many observers. It occurred to the writer that remarks founded upon a few objects which he had collected and mounted during a few months of the past and present year might stimulate others to make observations and collections of this description, and would be more acceptable to many than a technical or formal essay. It is often necessary to study objects according to a logical or scientific arrangement of them into groups, but the microscopist who keeps his eyes open will find subjects of investigation presented to him in defiance of all rules. This

is the case in reference to plants as much as to anything else; and the preceding illustrations may show how various species and parts may be studied, with the additional interest afforded by finding that the information they afford all tends to coalesce in certain broad philosophical ideas. Few persons can be botanists, in the sense of acquiring a complete knowledge of any considerable portion of the immense vegetable kingdom; but anyone in a few hours of recreation may acquire food for pleasant thought, and learn enough to appreciate many of the broad generalisations at which the greatest investigators arrive.