Owen, and which were by him referred to crustaceans probably resembling Limulus, were shown by the writer,

in 1862, to correspond precisely with those of the American Limulus (Polyphemus Occidentalis) (Fig.

5). I proved by experiment with the modern animal that the recurring series of groups of markings were produced by the toes of the large posterior thoracic feet, the irregular scratches seen in Protichnites lineatus by the ordinary feet, and the central furrow by the tail. It was also shown that when the Limulus uses its swimming-feet it produces impressions of the character of those named

FIG. 5.-Trail of a modern kingcrab, to illustrate imitations of plants sometimes named Bilobites.

FIG. 6.-Trail of Carboniferous crustacean (Rusichnites Acadicus), Nova Scotia, to illustrate supposed Algæ.

Climactichnites, from the same beds which afford Protichnites. The principal difference between Protichnites and their modern representatives is that the latter have two lateral furrows

produced by the sides of the caraрасе, which are wanting in the for

mer.

I subsequently applied the same explanation to several other ancient forms now known under the general name Bilobites (Figs. 6 and 7).*

The tuberculated impressions known as Phymatoderma and Caulerpites may, as Zeiller has shown, be made by the burrowing of the mole

FIG. 7.-Rusophycus (Rusichnites) Grenvillensis, an animal burrow of the Siluro-Cambrian, probably of a crustacean. a, Track connected with it.

cricket, and fine examples occurring in the Clinton formation of Canada are probably the work of Crustacea. It is probable, however, that some of the later forms referred to these genera are really Algae related to Caulerpa, or even branches of Conifers of the genus Brachyphyllum.

Nereites and Planulites are tracks and burrows of worms, with or without marks of setæ, and some of the

* The name Bilobites was originally proposed by De Kay for a bivalve shell (Conocardium). Its application to supposed Algae was an error, but this is of the less consequence, as these are not true plants but only animal trails.

markings referred to Palæochorda, Palæophycus, and Scolithus have their places here. Many examples highly illustrative of the manner of formation of the impressions are afforded by Canadian rocks (Fig. 8).

Branching forms referred to Licrophycus of Billings, and some of those referred to Buthotrephis, Hall, as well

FIG. 8.-Palæophycus Beverlyensis (Billings), a supposed Cambrian Fucoid, but probably an animal trail.

as radiating markings

referable to Scotolithus,

Gyrophyllites, and As

terophycus, are explained by the branching burrows of worms illustrated by Nathorst and the author. Astropolithon, a singular radiating marking of the Canadian Cambrian,* seems to be something organic, but of what nature is uncertain (Fig. 9).

Rhabdichnites and Eophyton belong to impressions explicable by the trails of drifting

sea-weeds, the tail-markings of Crustacea, and the ruts ploughed by bivalve mollusks, and occurring in the Silurian, Erian, and Carboniferous rocks. Among these are the singular bilobate forms described as Rusophycus by Hall, and which are probably burrows or resting-places of crustaceans. The tracks of such animals, when walking, are the jointed impressions known as Arthrophycus and Crusiana. I have shown by the mode of occurrence

* Supplement to "Acadian Geology."

t "Canadian Naturalist," 1864.

of these, and Nathorst has confirmed this conclusion by elaborate experiments on living animals, that these forms are really trails impressed on soft

sediments by animals and mostly by crustaceans.

I agree with Dr. Williamson* in believing that all or nearly all the forms referred to Crossochorda of Schimper are really animal impressions allied to Nereites, and due either to worms or, as Nathorst has shown to be possible, to small crustaceans. Many impressions of this kind occur in the Silurian beds of the Clinton series in Canada and New York, and are undoubtedly mere markings.

It is worthy of note that these markings strikingly resemble the socalled Eophyton, described by Torell from the Primordial of Sweden, and by Billings from that of Newfoundland; and which also occur abundantly in the Primordial of New Brunswick. After examining a series of these markings from Sweden shown to me by Mr. Carruthers in London, and also specimens from Newfoundland and a large number in situ at St. John, I am convinced that they cannot be plants, but must be markings of the nature of Rhabdichnites. This conclusion is based on the absence of carbonaceous matter, the intimate union of the markings with the surface of the stone,

"Tracks from Yoredale Rocks," "Manchester Literary and Philosophical Society," 1885.

their indefinite forms, their want of nodes or appendages, and their markings being always of such a nature as could be produced by scratches of a sharp instrument. Since, however, fishes are yet unknown in beds of this age, they may possibly be referred to the feet or spinous tails of swimming crustaceans. Salter has already suggested this origin for some scratches of somewhat different form found in the Primordial of Great Britain. He supposed them to have been the work of species of Hymenocaris. These marks may, however, indicate the existence of some free-swimming animals of the Primordial seas as yet unknown to us.

Three other suggestions merit consideration in this connection. One is that Algæ and also land-plants, drifting with tides or currents, often make the most remarkable and fantastic trails. A marking of this kind has been observed by Dr. G. M. Dawson to be produced by a drifted Laminaria, and in complexity it resembled the extraordinary Enigmichnus multiformis of Hitchcock from the Connecticut sandstones. Much more simple markings of this kind would suffice to give species of Eophyton. Another is furnished by a fact stated to the author by Prof. Morse, namely, that Lingulæ, when dislodged from their burrows, trail themselves over the bottom like worms, by means of their cirri. Colonies of these creatures, so abundant in the Primordial, may, when obliged to remove, have covered the surfaces of beds of mud with vermicular markings. The third is that the Rhabdichnite-markings resemble some of the grooves in Silurian rocks which have been referred to trails of Gasteropods, as, for instance, those from the Clinton group, described by Hall.

Another kind of markings not even organic, but altogether depending on physical causes, are the beautiful branching rill-marks produced by the oozing of water