have more recently named it, Nematophyton, as a primeval gymnosperm allied to those trees which Unger had described from the Erian of Thuringia, under the name Aporoxylon.* Later examples of more lax tissues from branches or young stems, and the elaborate examinations kindly undertaken for me by Professor Penhallow and

referred to in a note to this chapter, have induced me to modify this view, and to hold that the tissues of these singular trees, which seem to have existed from the be

* "Palaeontologie des Thuringer Waldes," 1856.

Figs. 2, 3, and 4 are drawn from nature by Prof. Penhallow, of McGill College.

ginning of the Silurian age and to have finally disappeared in the early Erian, are altogether distinct from any form of vegetation hitherto known, and are possibly survivors of that prototypal flora to which I have already referred. They are trees of large size, with a coaly bark and large spreading roots, having the surface of the stem smooth or irregularly ribbed, but with a nodose or jointed appearance. Internally, they show a tissue of long, cylindrical tubes, traversed by a complex network of horizontal tubes thinner walled and of smaller size. The tubes are arranged in concentric zones, which, if annual rings, would in some specimens indicate an age of one hundred and fifty years. There are also radiating spaces, which I was at first disposed to regard as true medullary rays, or which at least indicate a radiating arrangement of the tissue. They now seem to be spaces extending from the centre towards the circumference of the stem, and to have contained bundles of tubes gathered from the general tissue and extending outward perhaps to organs or appendages on the surface. Carruthers has suggested a resemblance to Algæ, and has even proposed to change the name to Nematophycus, or "thread-sea-weed"; but the resemblance is by no means clear, and it would be quite as reasonable to compare the tissue to that of some Fungi or Lichens, or even to suppose that a plant composed of cylindrical tubes has been penetrated by the mycelium or spawn of a dry-rot fungus. But the tissues are too constant and too manifestly connected with each other to justify this last supposition. That the plant grew on land I cannot doubt, from its mode of occurrence; that it was of durable and resisting character is shown by its state of preservation; and the structure of the seeds called Pachytheca, with their constant association with these trees, give countenance to the belief that they are the fruit of Nematophyton. Of the foliage or fronds of these strange plants we unfortunately know nothing. They seem, how

ever, to realise the idea of arboreal plants having structures akin to those of thallophytes, but with seeds so large and complex that they can scarcely be regarded as mere spores. They should perhaps constitute a separate class or order to which the name Nematodendreæ may be given, and of which Nematophyton will constitute one genus and Aporoxylon of Unger another.*

Another question arises as to the possible relation of these plants to other trees known by their external forms. The Protostigma of Lesquereux has already been referred to, and Claypole has described a tree from the Clinton group of the United States, with large ovate leaf-bases, to which he has given the name Glyptodendron. If the markings on these plants are really leaf-bases, they can scarcely have been connected with Nematophyton, because that tree shows no such surface-markings, though, as we have seen, it had bundles of tubes passing diagonally to the surface. These plants were more probably trees with an axis of barred vessels and thick, cellular bark, like the Lepidodendron of later periods, to be noticed in the sequel. Dr. Hicks has also described from the same series of beds which afforded the fragments of Nematophyton certain carbonised dichotomous stems, which he has named Berwynia. It is just possible that these plants may have belonged to the Nematodendreæ. The thick and dense coaly matter which they show resembles the bark of these trees, the longitudinal striation in some of them may represent the fibrous structure, and the lateral projections which have been compared to leaves or leaf-bases may correspond with the superficial eminences of Nematophyton, and the spirally arranged punctures which it shows on its surface. In this case I should be disposed to re

*See report by the author on "Erian Flora of Canada," 1871 and 1882, for full description of these fossils.

+"American Journal of Science," 1878.

gard the supposed stigmaria-like roots as really stems, and the supposed rootlets as short, spine-like rudimentary leaves. All such comparisons must, however, in the mean time be regarded as conjectural. We seem, however, to have here a type of tree very dissimilar to any even of the later Palæozoic age, which existed throughout the Silurian, and probably further back, which ceased to exist early in the Erian age, and before the appearance of the ordinary coniferous and lepidodendroid trees. May it not have been a survivor of an old arboreal flora extending back even to the Laurentian itself?

Multitudes of markings occurring on the surfaces of the older rocks have been referred to the Algæ or seaweeds, and indeed this group has been a sort of refuge for the destitute to which paleontologists have been accustomed to refer any anomalous or inexplicable form which, while probably organic, could not be definitely referred to the animal kingdom. There can be no question that some of these are truly marine plants; and that plants of this kind occur in formations older than those in which we first find land-plants, and that they have continued to inhabit the sea down to the present time. It is also true that the oldest of these Algae closely resemble in form plants of this kind still existing; and, since their simple cellular structures and soft tissues are scarcely ever preserved, their general forms are all that we can know, so that their exact resemblance to or difference from modern types can rarely be determined. For the same reasons it has proved difficult clearly to distinguish them from mere inorganic markings or the traces of animals, and the greatest divergence of opinion has occurred in recent times on these subjects, as any one can readily understand who consults the voluminous and well-illustrated memoirs of Nathorst, Williamson, Saporta, and Delgado.

The author of this work has given much attention to these remains, and has not been disposed to claim for the

*

vegetable kingdom so many of them as some of his contemporaries. The considerations which seem most important in making such distinctions are the following: 1. The presence or absence of carbonaceous matter. True Algæ not infrequently present at least a thin film of carbon representing their organic matter, and this is the more likely to occur in their case, as organic matters buried in marine deposits and not exposed to atmospheric oxidation are very likely to be preserved. 2. In the

absence of organic matter, the staining of the containing rock, the disappearance or deoxidation of its ferruginous colouring matter, or the presence of iron pyrite may indicate the removal of organic matter by decay. 3. When organic matter and indications of it are altogether absent, and form alone remains, we have to distinguish from Algæ, trails and burrows similar to those of aquatic animals, casts of shrinkage-cracks, water-marks, and rill-marks widely diffused over the surfaces of beds. 4. Markings depressed on the upper surfaces of beds, and filled with the material of the succeeding layer, are usually mere impressions. The cases of possible exceptions to this are very rare. On the contrary, there are not infrequently forms in relief on the surfaces of rocks which are not Algæ, but may be shallow burrows arched upward on top, or castings of worms thrown up upon the surface. Sometimes, however, they may have been left by denudation of the surrounding material, just as footprints on dry snow remain in relief after the surrounding loose material has been drifted away by the wind; the portion consolidated by pressure being better able to resist the denuding agency.

The footprints from the Potsdam sandstone in Canada, for which the name Protichnites was proposed by

* "Impressions and Footprints of Aquatic Animals," "American Journal of Science," 1873.