by the different shape of its post-dorsomedian plate, from C. cuspidatus by the different arrangement of the grooves on the outer surface of its cranial shield, and from both by the peculiar sculpture of its bony plates.

Cephalaspis Campbelltonensis. Sp. Nov.

Head shield (the only part known) large, somewhat pointed in front, obliquely rounded at the sides anteriorly, and produced behind into moderately elongated, slightly incurved cornua. Maximum breadth about seven inches. Orbits varying in outline from nearly circular to longitudinally broad ovate, sub-central, approximated, placed at distances from each other varying in different specimens from once to thrice the diameter of the orbit itself. Antorbital prominences rounded-conical; interorbital prominence also conical but somewhat elongated longitudinally; postorbital valley bounded by two narrow raised ridges, each of which starts from a prominence immediately behind the orbit: about halfway between the orbits and the posterior margin these ridges coalesce so as to form a single, broad and prominent but somewhat obscurely defined, posterior ridge.

Outer surface, which is very rarely preserved, polished and almost smooth to the naked eye. When examined under a lens it is seen to be minutely and densely pitted, the pits being very irregular in their shape, size and method of arrangement. Where the enamel is removed the surface is divided into numerous well marked polygonal areas.

Large fragments of the head-shield of this species are abundant in the Campbellton breccia, but the most perfect specimens yet obtained do not shew the outline of the posterior margin of the shield at all clearly. The orbits and the prominences and depressions in the central portion of the shield are often well defined, but the specimens are always crushed and nearly always exfoliated. Portions of the true outer layer of the test have been seen only on the central portion of the outer margin of the sides of one large fragment, and on the extremities of the cornua in two or three other specimens.

The genus Cephalaspis has been divided by E. Ray Lankester into three subgenera, viz., Eucephalaspis, Hemicyclaspis and Zenaspis, but as Hemicyclaspis is stated to be devoid of cornua it is clear that the C. Campbelltonensis cannot belong to this subgenus. Of the two which remain, Eucephalaspis and Zenas

pis have precisely similar head shields, but the body of Zenaspis has a dorsal scute placed immediately behind the posterior spine. In the absence of any knowledge of the body of the Campbellton species, therefore, it is uncertain to which of these two subgenera it should be referred.

Including the C. Dawsoni of Lankester, from Gaspé, all the specimens of Cephalaspis hitherto described are said to be characterized by a surface ornamented by raised tubercles, so that the C. Campbelltonensis may be readily distinguished by its minutely pitted sculpture. In general outline the head shield of the present species appears to be very much like that of the Eucephalaspis Powriei from the Old Red Sandstone of Forfar

shire.

Ctenacanthus latispinosus. Sp. Nov.

Compare C. ornatus, Agassiz.

Recherches sur les Poissons

Fossiles, Vol. 3, page 12, Table 2, figure 1.

Fin spines small (as compared with those of most of the other species of the genus) compressed laterally either elongated, slightly curved and tapering rapidly from a rather broad base to an obtuse point, or comparatively short, straight and triangular. Posterior margin somewhat concave, and bearing on its upper portion certainly one row and presumably two rows of short, conical hooklets, which curve obliquely downwards. Anterior margin thin, straight or gently convex, and unarmed. Surface marked on each side by from 15 to 20 longitudinal ribs, which swell out at regular intervals, of about one-third of a line apart, into subangular, equidistant nodes.

Length of the largest spine collected, about two inches and a half: maximum breadth of the same at the base, about three quarters of an inch.

The few spines of this species collected by Mr. Foord are all partly imbedded in the matrix, so that the grooving of the poste rior margin is hidden from view, and only one row of hooklets is exposed.

Homacanthus. Sp. Undt.

Compare H. arcuatus, Agassiz.

Poissons fossiles du Vieux

Grès Rouge, page 113, Table 33, figures 1-3.

Fin spine rather large (for a Homacanthus) compressed laterally, distinctly curved, slender, elongated and tapering very gradually from a narrow base to an apparently obtuse point. Upper

portion of the posterior margin armed with one or more rows of conical hooklets, which curve obliquely downwards. Surface ornamented by longitudinal ribs, with fine oblique striations.

Length about 17 lines, breadth at base about 3 lines.

Only one imperfect and badly preserved specimen was obtained, one side of which is buried in the matrix. It differs from the spines of Ctenacanthus latispinosus in its more slender proportions, more arcuate form, and apparently also in its surface ornamentation. As far as can be ascertained at present this spine appears to be very similar to the H. arcuatus of Agassiz, in almost every respect but that of size, the Campbellton species being much the larger of the two.

CRUSTACEA.

Pterygotus. Sp. Undet.

The occurrence of this genus at Campbellton is indicated by a fragment shewing the characteristic sculpture of semi-circular plicæ, and by a single ramus of the chela of an antenua, which must have belonged to a very large species. This ramus, which is not quite perfect at either extremity, is about two inches and a quarter in length, and nearly half an inch in breadth at its largest end. It bears on its inner margin four or five unequalsized, but comparatively large teeth, one of which is of much greater dimensions than the rest, -with a number of smaller ones between them. All the teeth are compressed and longitudinally striated most of them are ovate-lanceolate in outline, the basal portion being slightly constricted, but some of the small ones are simply conical.

MOLLUSCA.

Cyclora valvatiformis. Sp. Nov.

Shell very small, depressed turbinate, broader than high, spire much depressed: whorls three and a half, ventricose, rounded, increasing very rapidly in size, so that the greater part of the shell is formed by the last one, sutures deep; umbilicus between one-third and one-fourth of the diameter of the body whorl, deep in the centre and rounded at the margin. Mouth nearly circular but slightly angular posteriorly, next to the suture; lip thin and somewhat spreading. Surface nearly smooth, marked only by a few faint striæ of growth.

This specics was found in great abundance both by Mr. Foord and Principal Dawson, the most perfect specimens being those which were obtained from weathered surfaces. The resemblance of this little shell to the Cyclora minuta of Hall, from the Hudson River Group of Cincinnati, is certainly very close. The only differences that can be noticed between them at present, judging by Meek's detailed descriptions of Hall's species, are that the aperture of C. valvatiformis is slightly subangular behind and the lip somewhat expanded, whereas the phrases used to describe the corresponding parts of the shell of C. minuta are simply— aperture circular, lip thin." It is scarcely likely, however, that a shell which occurs associated with remains of Coccosteus and Cephalaspis, on the same small hand specimens of rock, is identical with a species from such a different geological horizon as the Hudson River Group.

Cyclora turbinata. Sp. Nov.

Shell very small, turbinate or turbinate conical, about onethird higher than broad, spire elevated: whorls four or four and a half, ventricose, rounded, increasing rather slowly in size, sutures rather deep: body whorl also rounded, base imperforate, aperture sub-circular, slightly angular behind: lip thin and somewhat spreading.

Length three lines, maximum breadth two lines.

Not more than about half a dozen specimens of this little shell have been collected. The species is invariably found associated with the C. valvatiformis, from which it differs in its much more elevated spire and closed umbilicus. Like C. valvatiformis, the present species is somewhat similar to one of the diminutive gasteropoda of the Hudson River Group of Ohio, the Cyclora parvula of Hall, but the body whorl of the latter shell is described as subangular, and its umbilicus as not quite closed.

NOTE ON A FERN ASSOCIATED WITH PLATEPHEMERA ANTIQUA, SCUdder.

BY J. W. DAWSON, LL.D., F.R.S.

The oldest remains of insects known to geologists, those of the Erian (Devonian) shales of St. John, New Brunswick, occur in beds rich in plant remains. It was indeed solely by means of the extensive quarrying operations carried on by Messrs. Hartt and Matthew in these beds in search of fossil plants, that the insect remains were discovered. In less thoroughly explored beds, fossils so rare and so obscure could not have been found. It is natural therefore that fossil plants should occur on the same slabs with the insects. On one of these, holding a fragment of the wing of Platephemera antiqua, there appears a con-siderable portion of a frond of Pecopteris (Aspidites) serrulata, Hartt, a common species in these beds, and also a small fragment of a leaf of the still more common Cordaites Robbii. It appears that Dr. Geinitz of Dresden saw this specimen in 1866, and not being at that time familiar with the ferns of the Devonian of New Brunswick, very naturally supposed that the frond was that of the closely allied P. plumosa of Brongniart,. and on this ground he was induced to hint a suspicion that the specimen was of Carboniferous age. Dr. Scudder referred to this opinion of Geinitz in his paper on Devonian insects in the Geological Magazine, Vol. V.; and gave reasons sustaining the Devonian age of both fern and insect. I did not think it necessary to refer publicly to the matter, but took occasion to explain the true state of the case in a private letter to Geinitz; and in my report on the Devonian plants of Canada I quoted Hartt's. description in full, and noticed the distinctness of his species. from P. plumosa.

I find, however, that this doubt has been revived by Dr. Hagen in a paper on Devonian insects in the Bulletin of the Museum of Comparative Zoology for the present year (Vol. viii. No. 14). Dr. Hagen does not profess to be an authority in fossil plants, but fortifies his statements by a letter from Mr. Lesquereux, which does not however touch the question at issue, as he does. not appear to have compared the specimen or Hartt's species with