Germar, of which I have not been able to see either figures or specimens that may possibly approach more nearly than those I have just noticed. For these reasons it only seems judicious to refrain from attaching a specific name to this species, it being at the least possible that it may ultimately prove to belong to one of these last-mentioned forms.

The present species has occurred in the Fish-bed, in the Grey Stone overlying, and in the Dun Stone in the New Quarry at Fulwell; and I have taken a single example from the Fish-bed in the Old Quarry of the same place.

Family 2. LEPIDOIDEI.

PALEONISCUS VARIANS, Kirkby. Pl. XVIII. fig. 2.
Annals of Nat. Hist. 3 ser. vol. ix. p. 267.

The maximum length of this Fish is from 3 to 4 inches. It usually occurs much less, often being only 13 inch long. Its maximum breadth is about an inch; but this is a point subject to much variation, the breadth of some specimens being one-third of the entire length, and of others only one-fifth. The body continues of similar width up to the dorsal and ventral fins; it then contracts somewhat rapidly to half the maximum width. The head varies in length from one-third to two-sevenths of the entire length, and its breadth is usually a little less than the greatest width of the body. The tail is moderately but decidedly heterocercal.

The fins are of median size. The pectoral, which is about half an inch long in mature examples, and placed at the junction of the ventral and median thirds of the body, consists of from 18 to 20 slender rays, with one or more short but strong spine-like rays in front. The ventrals are about inch shorter than the pectorals, and placed 13 inch from the snout; the rays are also less in number, though stouter, than those of the pectorals. The anal is 25 inches from the snout, or inch behind the ventrals; it numbers from 8 to 10 jointed rays, which are stronger and longer than those of the ventrals. The dorsal is placed about midway between the ventrals and the anal, or two inches from the snout; it is larger than the anal, being inch long, and it has from 10 to 12 jointed rays of similar strength to those of the anal. In advance of the longest rays of the anal and dorsal are two or three short, pointed, unarticulated rays; one or two similar rays are placed in front of the ventrals. The first segment of the articulate rays of the anal and dorsal is much longer than the succeeding segments-the succeeding segments being only about as long as wide. Each ray is covered by long unsegmented scales, which usually hide from view the raystructure. The front margin of each of these fins is protected by sharply pointed, linear fulcral scales. The caudal is of moderate size and slightly forked, spreading out gently from a base half as wide as the maximum breadth of the body to fully two-thirds of that breadth; it is formed of 17 or 18 rays, as strong as those of the dorsal and anal, and the rays of the ventral lobe are, of course, the

longest. The structure of the rays resembles that of the last-mentioned fins the first segment is of considerable length, being fully four times as long as those that succeed it; the succeeding segments are slightly shorter than the equivalent segments in the anal and dorsal, so as to be wider than long. They appear to bifurcate rarely; and they are covered with longitudinal scales, like those of the fins just mentioned. Each lobe is protected along its dorsal and ventral borders with comparatively large fulcral scales. The rays of the caudal, at least those of the central and ventral portions of it, are affixed to long interspinous bones, two rays being attached to each. The rays of the anal and dorsal fins are also attached to interspinous bones, there being in these instances, however, an ossicle to support each ray.

In well-preserved examples the head is seen to be bluntly coneshaped. The gape is comparatively small, and the orbit is large and oval. The opercular bones, which form the great bulk of the head, have a semicircular edge behind, and a surface with rugulose ornamentation. Below and behind the inferior maxillary are generally seen the branchiostegal rays, of which there appear to be nine.

The scales are large, comparatively thick, rhomboidal (variously modified), smooth to the eye, but finely shagreened when magnified, and with plain margins. The dorso-ventral series are arranged in sloping curves; there are 36 such series along each flank, and 14 or 15 scales in each series in the thoracic region. The lateral line is marked by a longitudinal series of notched scales, somewhat above the centre of the body; these scales are wider than any of the others on the flank; the scales above and below the lateral line decrease in size gradually, as they do also from the anterior portion of the body backward. Though the scales have been described as smooth to the eye, it should be mentioned that they are all marked, just within their exposed margins, with two or more lines of increment of great regularity. The overlap of the scales is considerable; their articulation is assisted also, as in other Palæonisci, &c., by teeth that project from the dorsal margin, and fit into sockets on the reverse of the ventral portion of the scales. In advance of each lobe of the caudal fin, and of the anal and dorsal, are placed four or five large scales, which are altogether different in form from the scales of the flank. They graduate from ovate or bluntly pointed scales to such as are lanceolate; and those of the latter form pass by insensible modifications into the fulcral scales or rays of the finborders.

There is more than one species to which this Palæoniscus has some resemblance; P. Voltzii, P. fultus, P. angustus, and P. Vratislaviensis, for instance, all appear related, though apparently distinct, forms. Perhaps the Fish that this species most nearly resembles is Palæoniscus glaphyrus, Ag., of the Marl-slate. It resembles it in size, largeness of scales, smallness of gape, size of orbit, and in position and structure of fins; but it differs from it in having a comparatively larger head-the head of P. glaphyrus being only one-fifth of the entire length, whereas that of P. varians reaches one-third-and

in having a large operculum, scales that vary more in size, and in having them with plain margins, while P. glaphyrus has them deeply serrated. P. glaphyrus, moreover, does not appear to have possessed the large notched scales that mark the lateral line in P. varians. With these differences before me, I have not hesitated to describe the present form as a distinct species.

P. varians has occurred in the laminated marls at the base of the Upper Limestone, in the strata immediately underlying the Fishbed, in the Dun Stone, and in the Main or Honeycomb Stone in the New Quarry at Fulwell; also in the Fish-bed in the Old Quarry, Fulwell. And the fragment of a Fish that I met with in the laminated limestone of Marsden apparently belongs to this species. PALEONISCUS ABBSII, Kirkby. Pl. XVIII. figs. 3a, 3b.

Annals of Nat. Hist. 3 ser. vol. ix. p. 268.

Length from snout to end of caudal fin rather more than 4 inches; length of body 3 inches. Greatest width, which is a little in advance of the ventral fin, inch; from this point the body contracts gradually to inch. From these measurements it will be seen that this species is an elongated form, the length of body being nearly six times the maximum width.

The head is an inch in length, and half an inch in breadth. The fins are of moderate size. The pectorals are about inch in length, and consist apparently of about 20 slender rays, with two short spine-like rays in front. The ventrals, which are comparatively small, are situated 17 inch from the snout; the anal is 2 inches from the snout; and the dorsal is placed midway between the ventrals and the anal, or 24 inches from the snout. Of the anal and dorsal, the latter is the larger; but none of the specimens found show either these fins or the ventrals in a condition that permits the rays to be counted. The caudal fin springs from a tail apparently more heterocercal than in P. varians; it consists of 15 or 16 stout rays, that are margined on each lobe by numerous pointed fulcral scales.

The scales resemble those of P. varians, except in being longer compared with their breadth. They are arranged in more inclined dorso-ventral series than in P. varians, and there are about 12 in each series; but the scales are generally so much dislocated and confused as to render their numeration a matter of difficulty.

The distinguishing feature of this Fish is its elongated form; and, except in being so much longer, it differs little from the preceding species.

From another elongated form of the same genus, P. longissimus, Ag., it differs in having much larger scales, and in their being smooth and unserrated at the margin; also in the relative position of the dorsal and ventral fins being different in the two species. From P. Kablike, another elongate Permian species, described by Dr. Geinitz, it differs in having a much blunter snout and smaller fins, as well as in a less elegant general form.

Examples, chiefly fragments of about half-a-dozen individuals, have occurred in the Fish-bed of the Old Quarry at Fulwell.

PALEONISCUS ALTUS, Kirkby. Pl. XVIII. fig. 1.

P. latus, Kirkby, Annals of Nat. Hist. 3 ser. vol. ix. p. 268*. Greatest length 28 inches; length of body 2 inches; maximum breadth rather more than 1 inch, or more than one-half of the length of the body; breadth of tail inch; head ginch long, and the same in width.

The general form of this fish is gibbous; the tail is only slightly heterocercal. The pectoral fins are small and slender; the ventrals, which are also small, are placed 11 inch from the snout, and the anal is 1 from the same point; the dorsal, which is the most robust of the fins mentioned, is situated at a point between the anal and the ventrals. The caudal is wide, and has about 24 strong rays: both lobes of this fin, as well as the anterior borders of the dorsal and anal, are protected by fulcral scales as in both the preceding species; and in advance of the fulcrals are the large " écailles impaires," which so generally accompany them.

66

The scales of the flanks are of the same type as those of P. varians, but are comparatively wider and shorter. They are robust, and smooth except in having marginal lines of growth, and are arranged in steeply sloping dorso-ventral series.

The bones of the head are ornamented with the same kind of rugulose sculpturing as that which characterizes P. varians. The orbit is also large.

I do not know of any species of Palæoniscus that approaches P. altus in its great relative width of body, although in general form of scales, in fin-structure, and in ornamentation of the bones of the head it differs little from the two preceding species.

One or two examples of this species have been found in the Fishbed of the Old Quarry, Fulwell.

PALEONISCUS ANGUSTUS?, Agassiz.

Along with the Palæonisci previously described has occurred a specimen of what appears to be another species. It is chiefly distinguished by its large fins and produced snout, and in general appearance has much resemblance to P. angustus, Ag., to which I provisionally refer it.

The specimen is in the possession of Mr. W. M. Wake, and was found in the Old Quarry, Fulwell.

PLANTE.

ULMANNIA SELAGINOIDES, King.

Several specimens have occurred, usually more or less fragmentary, of a vegetable that evidently belongs to the above species. The specimens are generally impressions, with traces of carbonaceous matter, the nervation of the leaves being occasionally shown on the impressed surface. The finest example, in the cabinet of Mr. Wake, is eleven inches long, with a branch of three and a half inches.

*At the suggestion of Sir Philip Egerton, I adopt the term altus for this species, latus having been already applied to another species of Palæoniscus.

ULMANNIA, sp.?

I have also examples of another Plant, belonging apparently to the same genus as the preceding, but they are not sufficiently perfect to allow me to describe from them.

CALAMITES ARENACEUS?, Brongn.

Among the vegetable remains are two specimens of a Calamite, which, on Mr. Howse's suggestion, I doubtfully refer to this species. The specimen in my possession is 4 inches in length and nearly an inch in width. It shows two joints, 11⁄2 inch distant from each other; and the impressed surface is rather finely striated longitudinally.

Besides the preceding species, a single specimen has occurred of a large reed-like Plant which I have not been able to identify. It is 21 inches long and 1 inch broad, rather coarsely striate longitudinally, and without any indications of constrictions or joints. The specimen is in the collection of Mr. Arthur Dixon.

I have to acknowledge the assistance of Mr. Howse in determining the Plants.

So far as may be judged from the fossils I have described, the physical conditions that prevailed during the deposition of the Fulwell Fish-bed were similar to those that obtained during the accumulation of the Marl-slate towards the commencement of the Permian era. The characteristic fossils of the Marl-slate are Fishes belonging to the genera Palæoniscus, Acrolepis, Pygopterus, Platysomus, and Coelacanthus, the species of the first genus being by far the most common. With the Fish occur the remains of Plants, chiefly belonging to Ulmannia selaginoides; and, rarely, examples of Lingula, Discina, and Myalina. The facies of this small fauna seems to me to be decidedly estuarine, though with a greater tendency to approach freshwater than marine conditions; for while the vegetable remains, which indicate terrestrial and freshwater conditions, are distributed generally throughout the whole of the Marl-slate, the Mollusca, which seem as certainly to indicate marine conditions, are confined to a very limited area in the same deposit. In the Fulwell Fish-bed we have Fish belonging to the same genera, and Plants belonging to the same species, as those that occur in the Marl-slate, besides other Plants whose occurrence there is not recorded. In this bed there are no Mollusca, nor is there, as I have before observed, a single trace of any marine organism. It would, therefore, seem as if the physical conditions of the Fulwell Bed had been even less marine than those of the Marl-slate; so that it is not unlikely that in its small group of species we see part of a freshwater fauna of the Permian period.

Another inference appears warrantable in respect to the Fish; that is, that the presence of so predaceous-looking a Fish as Acrolepis among small and comparatively harmless Palæonisci evidently indicates that the latter were pursued and preyed upon by it. The association merely of these Fish suffices to justify this inference; but