For the water of the contained air freezing on the sides of the boring (like hoar frost) would raise the temperature around the bulb, by the latent heat set free in freezing, to zero, C. And the humidity of the air being continually renewed, this process would always go on.

"A thermometer is, in short, incapable of taking the temperature of ice, unless that ice be dry."* These are the words of that eminent man Principal Forbes in the last (I think) of his papers on the subject of glaciers.

On a Friday evening like this, eight years ago, another eminent man, whose loss to science we all deplore, addressed the audience then assembled here on the subject on which I have now been addressing you. He possessed every accomplishment for the investigation of it, and had studied it long and successfully; he was versed in physics, and was an able and original mathematician. I speak of Mr. William Hopkins, whose name will always be held in honour by reason of his many contributions to the science of glaciers, and especially because of the explanation he first gave of the formation of crevasses. In the introductory part of the discourse which Mr. Hopkins then delivered he "insisted on the necessity of a more exact definition of terms and more accurate modes of mathematical reasoning than those which had up to that time characterized the discussion of glacial phenomena."

That, on the necessity of which he insisted and for which he laboured, I also have laboured for. I have sought to bring the discussion of glacial phenomena out of the wide region of scientific opinion, and place it in that of exact science. That work is, however, still far from being completed. Life, too, is short, and the power to pursue studies such as these is sometimes shorter in duration than life itself. But other workers are behind.

WEEKLY EVENING MEETING,

Friday, May 20, 1870.

[H. M.]

SIR HENRY HOLLAND, Bart. M.D. D.C.L. F.R.S. President, in the Chair.

WEEKLY EVENING MEETING,

Friday, May 27, 1870.

WARREN DE LA RUE, Esq. Ph.D. F.R.S. Vice-President,
in the Chair.

J. W. DAWSON, LL.D. F.R.S. &c.

PRINCIPAL OF M'GILL UNIVERSITY, MONTREAL,

On the Primitive Vegetation of the Earth.

THE existing flora of the earth is not primitive or primeval, but the last of a long series of dynasties of plants, extending from the oldest geological periods to the present time. Geology reveals to us several Tertiary floras, which perished before the modern period. These were preceded by the floras of the Mesozoic and these by those of the Palæozoic. Our attention will be directed this evening to the oldest of the Paleozoic floras, that of the Devonian and Upper Silurian periods.

To understand this old flora, it will be necessary for us to glance at the general divisions of the vegetable kingdom at present, and their range in geological time. Modern plants admit of a twofold division, into Phænogamous or flowering plants, and Cryptogamous or flowerless plants. The former have true seeds and fruits, the latter multiply by minute spores. Each of these great groups admits for our present purpose of a threefold division. Under Phænogams we have (1) Angiospermous Exogens, having stems with bark, wood and pith, and beginning their existence with two cotyledons, or seed leaves. These plants largely predominate among our trees, shrubs, and herbaceous plants, especially in the temperate climates. (2) Endogens, destitute of the divisions of the stem above mentioned, and beginning life with one cotyledon. The Palms belong to this group, and also the Grasses. (3) Gymnosperms, having exogenous stems, but with wood of more simple structure than in the first group, and having the organs of fructification of very imperfect character. This group includes the Pines and Cycads. Under Cryptogams we have also three divisions:-(1) Acrogens, the Ferns, Club-mosses, and Horse-tails. (2) Anophytes, the Mosses and their allies. (3) Thallophytes, the Lichens, Fungi, and Sea-weeds.

If we trace these forms of vegetation backward in geological time, we find that some of them appear to be older than others. The true Exogens, dominant in the modern time, continue to be so in the

Tertiary period; and in this the Endogens also exist as at present. As we pass backward in the Mesozoic, these two groups lose their importance, and we reach what has been called the age of Gymnosperms-Pines and Cycads being predominant. At the same time we lose from the other end of the scale the Anophytes. The Thallophytes scarcely at all appear as fossils, except in the case of a few Algæ. Thus in the Mesozoic flora the lowest group of the Phænogams and the highest group of the Cryptogams meet together, and share the world between them, almost to the exclusion of other groups.

Passing bakeward into the Paleozoic, the Cycads, which were plentiful in the Mesozoic, appear to fail us. We still find Conifers or Pine trees and other Gymnosperms, and we have a strange and remarkable development of the more humble Lycopodiaceous and Equisetaceous plants, the Club-mosses and Horse-tails, into arboreal forms. This condition of things exists in the Carboniferous period, in which our great beds of coal were formed, and extends backward into the oldest period in which we have a land flora. The Devonian thus belongs to the age of Acrogens, though it is not destitute of other forms.

My attention was first directed to these ancient plants by the discovery made many years ago by Sir W. E. Logan, that in the Gaspé sandstones, belonging to this group, there are not only drifted fragments of plants, but others rooted in the beds of clay or shale which evidently formed the soils in which they grew. On careful study the plants most abundant in these root-beds proved to belong to a type of vegetation previously unknown, and indicated by the restoration in Fig. 1. The lower portion consists of a cord-like rhizoma or rootstock, having a scalariform axis and a thick cellular bark. These rhizomata fill the root-beds with a confused mat of long cords, giving off numerous slender roots in every direction. In one case they have accumulated to a sufficient extent to form a bed of coal of one or two inches in thickness, the oldest known in America. The stems of these plants, to which the name Psilophyton princeps was given, are slender and branching, and in certain states of preservation are sparsely covered with short, stiff, spine-like leaves. In other states they are nearly smooth, with mere points instead of leaves. The young stems terminate in crozier-like circinate curls, like those of young ferns, and the fructification consists of oval sacs or spore-cases, borne gracefully on slender bending stalks. The fruit of these plants reminds the botanist of that of the modern Pillworts, but the stems are of more complete organization, and equal in this respect to those of modern Club-mosses. Such combinations of character are not unusual in fossil plants, and illustrate what seems to be a general law, viz. that in the older geological periods low forms of life were allowed to assume characteristics subsequently limited to higher grades of being.

The plants of the genus Psilophyton, of which there are several species, probably grew on swampy flats liable to inundation. They constitute the most characteristic and abundant members of the Lower Devonian flora, and appear to have been equally abundant in

Europe as in America, though when occurring in fragments they have often been mistaken for Algæ and for roots.

FIG. 1.

Psilophyton princeps, the oldest known plant of America, restored.

a, Fruit, natural size; b, stem, natural size; c, Scalariform tissue of the axis, highly magnified. In the restoration, one side is represented in vernation, and the other in fruit.

More distinctly allied to the modern Club-mosses were the Lepidodendra, Leptophlea, and similar plants of the Devonian, which may be defined to be gigantic arborescent Club-mosses or Lycopodiums. These also are widely diffused in the Devonian of all parts of the world; and some of the genera, especially Lepidodendron, are still more abundant

Leptophleum rhombicum-A Lycopodiaceous tree of the Devonian.
Portion of a branch, natural size.

Still another feature of this ancient vegetation was the occurrence of dense brakes of Calamites, plants which were exaggerations of the modern Equisetums, attaining to a diameter of several inches, and to a great height, and in some cases with strong and woody stems. The Calamites of the Devonian much resemble those of the Carboniferous, and two of the species are found also in the latter system. Probably allied to the Calamites were the beautiful Star-leaves of the genus Asterophyllites, which occur spread out on the Devonian shales of St. John, New Brunswick, as if prepared by a careful botanist.

The Ferns are among the most beautiful plants of the modern world in point of foliage, and they are very ancient in regard to geological time, making their appearance abundantly in the Middle