Sphenopteris Newberryi, but shows some features not seen in the figure of that plant given in the Pennsylvania reports. From a study of the isolated fragments, in which form alone I could get it, the plant shows the following features: The pinnules of the lower pinnæ have the form of those similarly placed in S. Newberry, but are proportionally narrower at the base, distinctly separate, and more obliquely placed. In ascending, the pinnules of the upper pinnæ are finally reduced to circular segments of the lamina of the pinnæ, and now if seen apart would be taken to belong to a different plant. These upper pinnæ are placed obliquely and alternately. They are ovate lanceolate, 12 mm. long and 6 broad at base; having the general shape and mode of incision shown in S. decipiens of the Pennsylvania reports. But unlike that, the termination of the pinnæ is prolonged into an acute point. Proceeding still higher on the frond, these pinnæ are reduced to semicircular lobes of the broad wing of the rachis which now forms the entire lamina of the upper part of the frond or compound pinna. The nerves are to a great extent masked by the thick leathery character of the leaflets, but, so far as mode out, are as follows: In the lower distinct pinnules there is a strong inidrib which disappears before reaching the extremity of the pinnule, and gives it at first sight the appearance of a Pecopteris. The side nerves spring very obliquely from the midrib, diverge very slowly from it, curving gently out to the margin, and fork once or twice, being quite distant from each other. In the rounded lobes of the upper pinnae, the nerves rise from the whole base of the lobe curving gently outward, and downward, while forking as before. Here the nervation resembles that of S. dilatata as figured in the Fossil Flora of Great Britain. The same nervation marks the extremity of the frond. The most characteristic feature is the rarity of the nerves, and if the plant should prove to be new, would justify the specific name varinervis.

16. Hymenophyllites spinosus Göpp.? Fragments were found on the Raleigh road with coal 5, of a plant showing the basal portion of several pinnæ, which seem to be identical with the above named plant. Not enough material was obtanied to identify it with certainty.

17. Sphenopteris macilenta Lindl. and Hutt. This plant seems to be abundant on the Raleigh road, associated with coal 5. Good specimens were obtained.

18. Equisetites, species? A single sheath, resembling that of an equisetites, was found on the Raleigh road with coal 5.

19. Asterophyllites acicularis Daws.? A specimen showing several whorls of leaves having the character of the above named plant, was obtained from coal 5 on the Raleigh road. Not enough is shown for positive identification.

20. Trigonocarpon triloculare Hildreth. A single nut was found at Sewell Station with coal 9. A nut of a different character from any figured or described to my knowledge, was found at the horizon of coal 5 at Quinnimont. It is perfectly smooth with no markings, is about 12 mm. long and 5 mm. wide, cylindrical in shape, and bluntly rounded at the ends, one of which is furnished with curved stem-like appendages, as if for attachment. Nuts are quite rare, only these two being found.

21. Megalopteris Hartii Andr. In my last visit to the plant locality at Sewell Station, which furnished me on my former visit the specimens of Megalopteris, and which is coal in 9, I procured a few additional specimens of this plant, among which, by comparison with the plates which Professor Andrews has had the kindness to send me, I recognized his M. Hartii. Of this plant I have one specimen having the ends of the two leaves at the summit of the frond, showing about six inches of their length. Another specimen shows the termination of a much smaller frond, with three leaves. Along with these leaves I find several of a small Megalopteris which may prove a different species although the nervation, so far as it can be made out in the obscure state of all the plants found here, seems to be very near that of M. Hartii. Of the small plant, no more than two leaves together have ever been found, and no specimen shows the point of junction of these. The small size seems to be a constant feature. Such leaves are about 6 cm. long and 1 cm. wide; they are narrowly elliptical in shape, with a rather more acute termination than that of M. Hartii. The midrib seems to have been large for a leaf of this size, and very prominent. It leaves a deep rectangular impression.

22. Megalopteris, species? This plant is the most common form found at Sewell Station. It differs from all the forms of Professor Andrews' plants figured in the decided acuteness of the leaves or pinnules; in which respect it is more like M. Dawsoni From this latter plant it differs in the more decided elliptical outline of the leaves, the more rapid narrowing of the leaves toward their extremities, and most of all, in the nervation, which so far as can be made out, is near that of M. Harti, being fine, from closely placed slender nerves, which fork near the base and again near their middle; apparently, higher up, than in M. Hartii. The nerves are nearly parallel in their course; and curve very slowly outward to meet the border of the leaf. I have one specimen which shows one entire leaflet, and the base of another, which diverges from the rachis at the base of the first, showing the ordinary alternate arrangement on a winged rachis, of the leaflets in plants of this genus. The entire leaflet of the specimen, is 14 cm. long, and 2 to 3 cm. wide. It is strongly

narrowed toward the base, giving it an oblanceolate shape. Near the extremity, it is rapidly narrowed to an acute point. It is most probably a new species, and if so, might receive the specific name Sewellensis.

23. Sphenopteris obtusiloba Brongt. Found rarely at Sewell Station, with coal No. 9.

24. Palæopteris Jacksoni Schimp. Cyclopteris Jacksoni Daws. Only one or two small fragments of this plant, were found at Sewell Station with coal 9.

25. Sphenophyllum antiquum Daws. A small fragment only

was found at Sewell Station with coal 9.

26. Odontopteris Neuropteroides Newb. Good specimens of this plant were obtained from the horizon of coal No 5 at Quinnimont, where with the variety of Alethopteris grandifolia, it forms the most abundant plant. Some pinnules show an obscure lobing not unlike some of the pinnules of Sphenopteris Lesquereuxii. Some scattered broad pinnules, with undulating borders, were found here which, from their appearance, would seem to have belonged to some part of this plant.

27. Calamites approximatus Schloth. Found at Sewell Station, with coal No. 9.

The plants above named, with the exception of the few got from Sewell Station previously, and mentioned in my former paper, were all obtained in my last visit to the New River region. As I stated before, this list can by no means be taken as exhaustive of the plants, even at the locality where they were gathered. In no case could I spend more than a couple of hours in collecting, and, having no tools, my collection was made by picking up fragments fallen from the disintegrated outcrop. This was the case with the coal on the Raleigh road, which furnished so many of the above specimens.

The following fossils were obtained from near the top of No. 1 of the transition beds at Quiunimont and consequently just from the base of the conglomerate series. They were kindly determined for me by Dr. J. J. Stevenson. He states that in most cases they were too badly preserved for specific determi

nation.

Invertebrate fossils from the base of the Conglomerate Series at Quinnimont.

1. Productus cora D'Orb.

2. Athyris sp.

3. Spirifera Leidyi N. & P.
4. Aviculopecten.

5. Lima.

6. Myalina.
7. Macrodon.
8. Lithophaga.
9. Chænomya?

10. Fenestella.

Of these Dr. Stevenson states that the Macrodon is a new species, which also occurs commonly in the middle portion of the

Umbral limestone in Monongalia Co., W. Va. He thinks that the Myalina and Chonomya are new species. Also that the Lithophaga is so near that referred with doubt to L. lingualis of Phillips, by M. & W., that he cannot distinguish it, although this is a species of the St. Louis group.

I may state here that the Palaropteris Jacksoni of the Conglomerate series, is the typical plant, and very different from the plant found at Lewis Tunnel, and given in my previous paper as P. Jacksoni. I have additional specimens from Lewis Tunnel, which show without doubt that, as Professor Andrews has suggested, this latter is a new species.

It will be seen from the above, that the representatives of the Devonian flora of Canada are quite common in the Conglomerate Series. The uppermost strata of the Devonian in West Virginia are at least 3500 feet below coal 5; and still farther below coal 9, which affords the plants of most decided Devonian type. Besides, the forms of Megalopteris, the Palæopteris, and Sphenophyllum, we have probably Asterophyllites acicularis, and Cordaites Robbii, identical with the Devonian plants. The Cordaites has the nervation and termination of the leaves of C. Robbi, but I mark it doubtful, as I have no entire leaves. The Sphenopteris adiantorides, is a good deal like the plant figured as Cyclopteris obtusa, by Dawson in the Acadian Geology, although smaller. The plant marked doubtfully Alethopteris Serli, is very near A. discrepans. The Sphenopteris allied to S. Newberryi, in its upper pinnæ shows the mode of lobing, and has something of the aspect of S. marginata.

It will be noted that along with these plants we have some of the forms found in coal No. 1 of Ohio. The upper pinnæ of the plant identified with Sphenopteris macilenta I cannot distinguish from Dawson's Cyclopteris valida. Besides these, we have plants first found in Felling Colliery, England, associated with the "Low Main coal seam."

ART. XLIX.-Mineralogical Notes; by EDWARD S. DANA. No. III.-On new twins of Staurolite and Pyrrhotite.

1. ON STAUROLITE CRYSTALS FROM FANNIN CO., GEORGIA.

THROUGH the instrumentality of Prof. F. H. Bradley a large number of staurolite crystals have been recently received in New Haven, some of which show forms which are new and interesting. Prof. Bradley mentions two distinct localities, visited by him, which afford the staurolite in considerable quantities. The first is at Valley River, near Murphy, Cherokee Co., North Carolina. The crystals at this place are large

and coarse. They occur in a metamorphic rock which Prof. Bradley states belongs to the Cincinnati group.

The second locality is in Fannin Co., Georgia, 12 miles southeast of Ducktown, Tenn. The crystals obtained here are of quite uniform size, averaging an inch in length, and are very perfectly developed, the faces being smooth, though usually without polish. They occur in a soft mica schist, belonging geologically, according to the same authority, at the base of the Quebec Group. The rock is extensively decomposed so that the crystals are found abundantly loose on the surface of the ground. Large quantities of them may thus be picked up, of which perhaps one-tenth are perfect crystals. They are in general nearly free from the gangue.

The crystals are with rare exceptions twins. The most common twins are those well known in this species; that is, (a), those having the composition face -, with the vertical axes nearly at right angles to each other; and (b), those having the composition-face, with the vertical axes crossing at a angle of about 60°. Under these two types there is a very wide diversity of form arising from the extension or partial suppression of various of the occurring planes. A very common feature in the twins of the first kind mentioned is the almost entire suppression of one pair of the macrodome planes 1-7, opposite each other, with the corresponding extension of the other pair; this same kind of hemihedral development is also extended to the prismatic planes giving rise to forms of very oblique appearance.

Three new forms are shown in figures 1, 2 and 3. Figure 1 exhibits a new method of twinning not before observed in this

species, and making the third type of penetration twins peculiar to staurolite. The measured angle of the two brachypinacoids (-) for the crystals in their twinning position is 70° 30'. This gives for the composition-face i-3, which requires for the above angle 70° 18'. In this case the axis of revolution very nearly coincides with the plane and the -, of the second individual is very nearly parallel to the prism I of the first. Attention has been called, by Prof. Dana, to the fact that if the