of both male and female flowers, in the development and structure of the embryo, and especially in their remarkable habit and sensible properties. I have already explained in sufficient detail their close agreement in floral structure, which warrants us in considering them as forming a single truly natural Order, not (as some authors think) divisible into several sections, each of equal value to Gnetaceae and Cycadeae.

In order to explain the relationship of Cycadeae to Conifers and Gnetaceae, and the position of the class Gymnosperms in the Vegetable Kingdom, it is necessary to say a few words on the morphology of Cycads.

On the simple axis or caudex of the female Cycas there are only four formations of leaves: 1st, The cotyledons, to which succeed (2nd) scale-like leaves, sheathing and protecting the others in bud; 3rd, True (pinnate) leaves, fitted for the assimilation of the sap; and, 4th, carpellary leaves. As each of these discharges an important vital function, they may all be regarded as essential. After the cotyledons, the tree, for a long succession of years, produces only scales and true leaves, till at last, when mature, it proceeds to flower, and, after a copious development of scales, produces verticils of carpels.

The spadix (or carpel) has the form of a somewhat contracted leaf, bearing towards the apex leaflets but slightly modified. Below, in place of leaflets, it bears ovules. We have here the earliest stage of metamorphosis, and the least altered type of the carpel. It is, indeed, impossible to imagine a less complex case of metamorphosis than one in which some parts of a foliar organ are expanded into organs adapted for assimilation, while others are contracted into reproductive organs.

In all other Phaenogams (not even excepting the male Cycas), when the leaves are changed into reproductive organs, whether stamens or carpels, their foliar nature is altogether modified, and the vis generatrix of the axis of growth is so much diminished and impaired that it loses its power of continued development, and can no longer produce new leaves: it is therefore terminated by a flower. In the female Cycas, on the other hand, the metamorphosis acting on the leaflets only, and not affecting the form and structure of the leaf, does not interfere with the vitality and vigour of the axis, which therefore grows out beyond the flower, and from its centre produces leaf-scales and a cluster of true leaves, so that the same axis produces repeatedly, in alternate succession, flowers and leaves.

In the male Cycas and other Cycadeae, as is usual in Phaenogams, the axis is terminated by the production of a flower. Both stamens and carpels are therefore borne on secondary axes developed after the expansion of the leaves, from which they are quite different in appearance, while the primary axis produces only scales and leaves and is continued indefinitely.

We have, therefore, in Cycads a very simple and unexpected link between Ferns and Phaenogams. In many ferns (Blechnum for instance), sterile and fertile fronds are produced in alternate succession from the same axis, just as in the female Cycas. In all other Cycadeae the flowers terminate the axis.

In the Cycadeae we find the archetype from which the whole vast series of Phaenogamous plants has been produced. In this Order we find not only the earliest conditions of leaf metamorphosis in the stamens which are of the least altered state, and the simplest forms of carpels, but also the prototypes of the different forms and relative arrangements of the axis and the leaf in Phaenogams. At one time all the essential forms of the leaf are produced on a simple axis, so as to give us the plan of a uniaxial plant; at another time one of the leaf forms (the stamen or the carpophyll), being produced on a secondary axis, we have the simplest condition of a pluri-axial plant. Of all living plants, therefore (excepting Cryptogams), Cycads have the best title to the appellation of "imagines Plantarum," (Urpflanzen or Idealpflanzen).

Conifers are very different in structure. For, whereas the reproductive organs are, in Cycads, always modified leaves, in Conifers axis and leaves are both made use of, but differently in the two sexes. In the male the stamens are altered leaves, and the axis is undeveloped, (definite). In the female the leaves are little altered from their ordinary condition and serve as bracts, while the axis springing from their axil is wholly, or at its apex, changed into an ovule, and thus becomes a reproductive organ. The structure of Gnetaceae is the

same.

Conifers and Gnetaceae are therefore directly opposed in the structure of the female flower to Cycads, and to most (or all?) Phaenogams. It may appear surprising that the mode of production of the ovule should be so variable, but it must be observed that there is a similar variation in Cryptogams. It is certain that in Lycopodiaceae, the sporangium, which is undoubtedly the analogue of the ovule, springs from the axil of a leaf and has the character of an

axillary bud.* In Ferns, on the other hand, it arises from the leaf or frond. Lycopodiaceae, therefore, as regards the sporangium or ovule, are the analogues of Conifers, and Ferns of Cycadeae.

As soon as this analogy between Gymnosperms and Cryptogams has been recognized, other indications of relationship at once attract our attention. Without discussing this question at length, which I am unwilling to do, lest it should carry me too far into fanciful analogies, I may remark that there is evidently the same degree of affinity between Lycopodiaceae and Ferns, as between Conifers and Cycads. For Lycopodiaceae agree with Conifers not only in the above mentioned point of the sporangia, but also in the modified condition of the leaves in the axils of which they are produced, and in their arrangement in aments, in their very copious branches, in the shape of the leaves, and in their whole habit, while Cycads resemble Ferns in the simple axis, in the shape and venation of the fronds, and also, as we have explained above, in the succession of the different leaf formations.

We thus see that in these Orders of plants the primary organs, that is the axis and leaves, are altered by metamorphosis in a totally different manner. In Lycopodiaceae and Coniferae the leaves are simple and do not vary much in size and shape, the axis is highly developed and much branched; the younger branches covered with green bark and pierced by stomates, discharge equally with the leaves the functions of vegetation and assimilation. Sometimes the leaves are much reduced in size, and the axis dilated into a leaflike form (Phyllocladus), alone discharges these functions. The axis also transformed by a powerful metamorphosis into an ovule or sporangium, is made use of as a reproductive organ,. In Cycads and Ferns, the axis is for the most part simple and of slow growth, completely covered with leaves and scales, often very short and subterranean. Thus the whole force of development is transferred to the leaves, which are of large size, wonderfully various in form, divided and articulated, and, inasmuch as sporangia and ovules are

* Hoffmeister, Vergl. Unters. 119. In a note on the sporangium of Selaginella, overlooking this earlier view, Hoffmeister adopts Mohl's opinion that the sporangium is a part of the leaf above which it originates, but as he has elsewhere said that this view can hardly be reconciled with nature, and that the development of the sporangium shows (beyond a doubt, is the expression in the paper quoted) that it is a bud in the axil of the leaf below it, even when it is inserted on the leaf above the axil, I do not well understand why this admirable observer has abandoned his first view which appears confirmed by careful observation.

produced on them, they discharge the functions both of vegetation and reproduction.

I add a few words on the place of Gymnosperms in the system. Cycads and Conifers are, in floral structure, more imperfect than any Order of Angiosperms. Their flowers, destitute of perigonium and ovary, consist of an ovule only, or of an ament formed of stamens or open carpellary leaves. In Gnetaceae the male and female organs are both covered by a perianth, but the ovary which is present in all Angiosperms (except perhaps Rhizanths) is wanting. The three Orders of Gymnosperms are, however, shown to be more perfect than any Cryptogamous Order, by the presence of true stamens and by the nature of the ovule. It is generally admitted that they occupy an intermediate position between Cryptogams and Angiosperms, both in the mode of production of the embryo and in the growth of the pollen, and that they form a passage from the one to the other. It is, I believe, no less certain that Cycads in the structure and germination of their embryo combine the characters of Dicotyledons and Monocotyledons and connect them with Isoeteae, and that in the venation of their leaves and the structure of their vegetative and reproductive organs they connect Angiosperms with Cryptogams. In the time at which they appeared on our earth, too, they are intermediate between Cryptogams and Angiosperms. I therefore think that Gymnosperms are a truly natural class, the place of which is between Cryptogams and Angiosperms.

XLII.—PROCEEDINGS OF THE SCIENTIFIC SOCIETIES OF LONDON.

1. ETHNOLOGICAL SOCIETY, (4, St. Martin's Place).

December 22nd.

MR. GALTON, F.R.S., read a paper on the "Domestication of Animals," which he introduced with the following observations:-" The domestication of animals is one of the few relics of the past whence we may justly speculate on man's social condition in very ancient times. We know that the domestication of every member of our

existing stock was originated in pre-historic ages, and, therefore, that our remote ancestors had accomplished, in a variety of cases, what we have been unable to effect in any single instance. The object of my paper is to discuss the character of ancient civilization, as indicated by so great an achievement. Was there a golden age of advanced enlightenment? Have extraordinary geniuses arisen who severally taught their cotemporaries to tame and domesticate the dog, the ox, the sheep, the hog, the fowl, the camel, the llama, the reindeer, and the rest? Or again, is it possible that the instincts of savages, combined with the qualities of the animals in question, may have sufficed to originate every instance of established domestication? It is to be presumed, in the first place, that animals would be originally domesticated in lands where they abounded in a wild state, and where the natives were skilled in capturing them. Unless the animals were easily obtainable we could hardly expect a sufficient number of experiments to have been made to yield a successful result. If they had been rare in all places and at all times, they would ipso facto be disqualified for domestication; for animals must be hardy and able to multiply freely under varying circumstances, else they would be of no importance as a domestic breed. Secondly, it is a fact familiar to all travellers, that savages frequently capture young animals of various kinds, and rear them as favourites. Human nature is generally akin: savages may be brutal, but they are not on that account devoid of our taste for taming and caressing young animals; nay, it is not improbable they may occasionally possess it in a more marked degree than ourselves, because it is a childish taste with us; and the motives of an adult barbarian are very similar to those of a civilized child." After giving various instances of the attachment of native tribes to pet animals, as detailed by travellers, Mr. Galton continued:-" It will be found on enquiry that few travellers have failed altogether to observe instances of wild animals being nurtured in the encampments of savages. If we consider the small number of encampments they visit on their line of march, compared with the vast number that are spread over the whole area, which is or has been inhabited by savages, we may obtain some idea of the thousands of places at which half-unconscious attempts at domestication are being made in each year. These thousands must themselves be multiplied many thousand-fold, if we endeavour to calculate the number of similar attempts that have been made since men like ourselves began to inhabit the world. I conclude from what I have stated that there