a similar observation this year in a Wisteria sinensis. Plants on my grounds have made an unusual second flowering. There were more blossoms in July than in April. Among them is a snow white variety, which has flowered annually for six years past at least. At this second flowering it took a notion to flower blue, not quite as deep a blue as the regular tint of the well known kind; but still anything but the white we have always had before. It was very difficult for my gardener to believe that in some way or another "some hybridization" had not been going on. Potatoes frequently change this way in the color of the tubers, when the intelligent farmer is sure "there must have been some mixing of the pollen which in some way affected the circulation and changed the color." Dahlias, chrysanthemums, balsams, and many other things with parti-colored flowers, frequently have some wholly of one of the mixed colors; but all this in some way is supposed to be the work of art.

These natural variations I regard with much interest as teaching us that the law of evolution is not wholly through seed, and that those botanists who look for it in the embryology of the reproductive organs are not wholly on the right track.

Physiologists usually commence their treatises with "the seeds;" as if the seed was the primary element in the organization of vegetation, instead of the final result. Not that they really teach it, but this order of treating the subject gives the public mind that impression. Mr. Darwin's ideas seem to arise from some such reasoning as this. It seems hardly possible to conceive of first existences from eggs or seeds. True we see most of the changes through this medium now; but if we find cases in abundance (and I think we might if we looked for them) like these of Trillium and Wisteria, where changes occur independently of sexual influence, they will at least suggest another law to account for the origin of species.

THE PRIMITIVE VEGETATION OF THE EARTH.

BY J. W. DAWSON, LL. D.

TWENTY years ago scarcely anything was known, even to those engaged in the study of vegetable fossils, of a land flora older than the great coal formation. In 1860, Goeppert, in his Memoir on the plants of the Silurian, Devonian, and Lower Carboniferous, mentions only one land plant, and this of doubtful character, in the Lower Devonian. In the Middle Devonian he knew but one species; in the Upper Devonian he enumerated fifty-seven. Most of these were European, but he included also such American species as were known to him. The paper of the writer on the Land Plants of Gaspé was published in 1859, but had not reached Goeppert at the time when his memoir was written. This, with some other descriptions of American Devonian plants not in his possession, might have added ten or twelve species, some of them Lower Devonian, to his list. In the ten years from 1860 to the present time, the writer has been able to raise the Devonian flora of Eastern North America to one hundred and twenty-one species, and reckoning those of Europe at half that number, we now have at least one hundred and eighty species of land plants from the Devonian, besides a few from the Upper Silurian. We thus have pre

sented to our view a flora older than that of the Carboniferous period, and, in many respects, distinct from it; and in connection with which many interesting geological and botanical questions arise.

Geologists are aware that in passing backward in geological time from the modern to the Paleozoic period, we lose, as dominant members of the vegetable kingdom, first, the higher phænogamous plants, whether exogenous or endogenous; and that, in the Mesozoic period, the Acrogens, or

higher cryptogams, represented by Ferns, Club-mosses, and Equiseta, share the world with the Gymnosperms, represented by the pines and Cycads, while the higher phænogams on the one hand, and lower cryptogams on the other, are excluded. Hence, the Mesozoic age has been called that of Gymnosperms, while the Paleozoic is that of Acrogens. These names are not, however, absolutely accurate, as we shall see that one of the highest forms of modern vegetation can be traced back into the Devonian; though the terms are undoubtedly useful, as indicating the prevalence of the types above mentioned, in a degree not now observed, and a corresponding rarity of those forms which constitute our prevalent modern vegetation.

It is my present object shortly to sketch the more recent facts of Devonian and Upper Silurian Botany, and to refer to a few of the general truths which they teach. The rocks called Devonian in Europe being on the horizon of the Erie division of the American geologists, which are much more fully developed than their representatives on the Eastern Continent, I shall use the term Erian as equivalent to Devonian, understanding by both that long and important geological age intervening between the close of the Upper Silurian and the beginning of the Carboniferous.

Just as in Europe the rocks of this period present a twofold aspect, being in some places of the character of a deposit of "Old Red Sandstone," and in others indicating deeper water, or more properly marine conditions, so in America, on a greater scale, they have two characters of development. In the great and typical Erian area, extending for seven hundred miles to the westward of the Apalachian chain of mountains, these rocks, sometimes attaining to a thickness of fifteen thousand feet, include extensive marine deposits; and except in their north-eastern border are not rich in fossil plants. In the smaller north-eastern area, on the other hand, lying to the eastward of the Apalachian range, they consist wholly of sandstones and shales,

and are rich in plant remains while poor in marine fossils. Hence it is the Devonian of Gaspé, of New Brunswick, and of Maine, with that of eastern New York, which have chiefly afforded the plants to be described below; and it is exclusively in these areas that we find underclays with roots, or true fossil soils. Most of the localities of fossil plants in the districts above mentioned have been visited, and their plants studied in situ by the writer. The Gaspé sandstones were first studied and carefully measured and mapped by Sir W. E. Logan. The Devonian beds of St. John's, New Brunswick, have been thoroughly examined and illustrated by Professor Hartt and Mr. Matthews, and those of Perry by Professor Jackson, Professor Rogers and Mr. Hitchcock. Professor Hall, of the Survey of New York, has kindly communicated to me the plants found in that State, and Professor Newberry has contributed some facts and specimens illustrative of those of Ohio.

In the Sandstone cliffs of Gaspé Bay, Sir W. E. Logan recognized in 1843 the presence of great numbers of apparent roots in some of the shales and fine sandstones. These roots had evidently penetrated the beds in a living state, so that the root-beds were true fossil soils, which, after supporting vegetation, became submerged and covered with new beds of sediment. This must have occurred again and again in the process of the formation of the four thousand feet of Gaspé sandstone. The true nature of the plants of these fossil soils I had subsequently good opportunities of investigating, and the most important results, in the discovery of the plants of my genus Psilophyton, are embodied in the restoration of P. princeps. This remarkable plant, the oldest land plant known in America, since it extends through the Upper Silurian as well as the Devonian, presents a creeping horizontal rhizome or root-stock, from the upper side of which were given off slender branching stems, sometimes bearing rudimentary leaves, and crowned when mature, with groups of gracefully nodding oval spore

cases.

The root-stocks must in many cases have matted the soils in which they grew into a dense mass of vegetable matter, and in some places they accumulated to a sufficient extent to form layers of coaly matter, one of which on the south side of Gaspé Bay is as much as three inches in thickness, and is the oldest coal known in America. More usually the root-beds consist of hardened clay or fine sandstone filled with complicated net-work or with parallel bands of rhizomes more or less flattened and in various states of preservation. In all probability these beds were originally swampy soils. From the surface of such a root-bed there arose into the air countless numbers of slender but somewhat woody stems, forming a dense mass of vegetation three or four feet in height. The stems, when young or barren, were more or less sparsely clothed with thick, short, pointed leaves, which, from the manner in which they penetrate the stone, must have been very rigid. At their extremities the stems were divided into slender branches, and these when young were curled in a crosier-like or circinate manner. When mature they bore at the ends of small branchlets pairs of oval sacs or spore-cases. The rhizomes when well preserved show minute markings, apparently indicating hairs or scales, and also round areoles with central spots, like those of Stigmaria, but not regularly arranged. These curious plants are unlike anything in the actual world. I have compared their fructification with that of the Pilulariæ or Pillworts, a comparison which has also occurred to Dr. Hooker. On the other hand, this fructification is borne in a totally different manner from that of Pilularia, and in this respect rather resembles some ferns; and the young stems by themselves would be referred without hesitation to Lycopodiaceæ. In short, Psilophyton is a generalized plant, presenting characters not combined in the modern world, and, perhaps illustrating what seems to be a general law of creation, that in the earlier periods low forms assumed characteristics subsequently confined to higher grades of being.