1

care to let no iron instrument come near them, have broken them by rapping them together. Then taking only the interior parts of these nodules I have pulverized them in a porcelain mortar. The magnetic particles were afterwards extracted by a magnet covered with paper. Now, placing these particles on a glass slide under the microscope, and adding the sulphate of copper solution, there was in a few moments a deposit of copper on several small perfect spherules, varying in size from the 100 to the 6 of an inch in diameter. I have placed some of these spherules under the microscope and now show them to the Society. It will be noticed that on one the copper is not deposited all over the sphere, but in ramified spider-like lines. On the cut surface of a meteorite, from Professor Sir Wyville Thomson's collection, which I also exhibit, the copper is precipitated in precisely the same manner as on the little sphere from the manganese nodule. Besides the spherules on which the copper is deposited, there are others generally of a larger size and dark colour. These are, so far as microscopic examination shows, quite like the particles on the mammillated outer surface of this Cape meteorite, also from Sir Wyville's collection.

These spherules have hitherto only been noticed in those deposits in deep water far from land, and where for many reasons we believe the rate of formation of deposits to be very slow.

They occur also only in those manganese nodules which come from the same deep-sea clays or deposits far from land.

The particles of native iron found in pumice-stones are not numerous, and never take the form of spherules so far as observed. Some of these particles of native iron may then come from the dredge. Other particles come from the pumice and the volcanic materials. Professor Andrews long since showed that minute particles of native iron existed in basalt and other rocks. And lastly the spherules of which I have been speaking appear to have a cosmic origin.

The reason, for these spherules occurring only in deposits far from land and in deep water, may be more apparent by reference to the annexed diagram, which might represent a section from the west coast of South America out into the Pacific 500 miles. Along the shores of the continent as at a we have an accumulation of river and coast detritus. At b in depths from 1400

VOL. IX.

2 M

to 2200 fathoms we have a globigerina ooze mostly made up of surface shells. At c, in a depth of 2300 to 3000 fathoms, all the surface shells are removed from the bottom. No coast detritus reaches this area, and we find in the deposit pumice stones, some volcanic ashes, manganese nodules, sharks' teeth, and ear bones of whales. It is only in areas like this that we find sharks' teeth and ear bones of cetaceans in any numbers. Some of them from the same haul are deeply surrounded with manganese deposit, and contain little animal matter; while others have no deposit on them, and seem quite recent. These, and other facts which might be mentioned, all argue for an exceedingly slow rate of deposition. Now it is in these same areas that the spherules of native iron and other magnetic spherules are found, both in the deposits and in the manganese nodules from them.

Finding them in this situation favours the idea that they are of cosmic origin, for in such places they are least likely to be covered up or washed away. It is certainly difficult to understand why the spherules on which the copper is precipitated have not become oxidised. If nickel be present in them, this may retard oxidation to some extent.

The manganese depositions in our ocean deposits are very different in structure and composition from any of the ores of manganese I have had an opportunity of examining, and the deposits of the deep sea far from land have not, so far as I know, any equivalents in the geological series of rocks.

All the subjects treated of in this paper are still under investigation, and at some future time I hope to present a much more detailed account.

These observations seem to me to give ground for the following conclusions:

First, That volcanic debris, either in the form of pumice stones, ashes, or ejected fragments, are universally distributed in ocean deposits.

Second, That pumice stones are continually being carried into the sea by rivers and rains, and are constantly floating on the surface of the ocean far from land.

Third, That the clayey matter in deposits far from land is principally derived from the decomposition of the feldspar in fragmental volcanic rocks, though in the trade wind region of the North Atlantic the dust of the Sahara contributes much material for clay.

Fourth, That the red earth of Bermuda, Bahamas, Jamaica, and other limestone countries, is most probably originally derived from the decomposition of pumice stone, while these limestones were in the process of formation.

Fifth, That the peroxide of manganese is probably a secondary product of the decomposition of the volcanic rocks and minerals present in the areas where the nodules of manganese are found.

Sixth, That there are many minute particles of native iron in deposits far from land; that some of these particles are little spherules; that these last, as well as some other spherules which are magnetic, have probably a cosmic origin.

Seventh, That the peroxide of manganese depositions in the deep sea are different in structure and composition from known cres of manganese.

Eighth. That we do not appear to have equivalents of the rocks, now forming in the deep sea far from land, in the geological series.

In conclusion, I have to acknowledge much assistance in these investigations from all my colleagues, especially my indebtedness to Sir Wyville Thomson and Mr Buchanan.

Since my return I have received many hints from Professors Tait, Geikie, Turner, Dr Purves, Mr Morrison, and other gentlemen.

In much of the mechanical work which an examination of these deposits has entailed, I have, both during the cruise and since my return, had the assistance of Frederick Pearcey.

4. On New and Little-known Fossil Fishes from the Edinburgh District, No. I. By R. H. Traquair, M.D., F.G.S.

FAMILY PALEONISCIDE.

GENUS Nematoptychius, Traquair, 1875.

This genus was instituted by the author for the reception of the Pygopterus Greenockii of Agassiz, and characterised in the "Annals and Magazine of Natural History" for April 1875. It differs from Pygopterus in the form of the scales of the flank, which are much higher than broad, and having their articular spine arising from the whole, or nearly the whole of the upper margin; in the structure of the pectoral fin, in which all the rays are articulated for the greater part of their extent; and in the form of the anal, which is in shape like the dorsal, and is not produced backwards in the peculiar fringe-like manner characteristic of Pygopterus.

Since the publication of the notice above referred to, remains of N. Greenockii have turned up in two other localities near Edinburgh, viz., near Juniper Green in the horizon of the Wardie Shales (Museum of Science and Art), and at Raw Camps, near Mid Calder, in that of the Burdiehouse Limestone (Collection of the Geological Survey of Scotland).

The following is new to science.

Nematoptychius gracilis, sp. nov. Traquair.

The

Of this two specimens have occurred at Gilmerton. more perfect of these, compressed on its side, displays the entire contour of the fish, including all the fins, and measures 9 inches in total length, by 1 inch in depth between the head and the ventrals; the length of the head is contained about 4 times in the total. The ventral fins arise opposite a point 3 inches distant from the tip of the snout; the commencement of the anal is midway between that of the ventral and of the caudal; the dorsal is situated nearly opposite the anal, commencing only inch in front of it. The form of the fish is thus elongated and slender, gradually tapering from behind the shoulder towards the tail, the dorsal fin being situated very far back. The other and slightly longer spe

cimen lies compressed on its back, and shows the under surface of the head, both pectoral and both ventral fins, with traces also of the dorsal, anal, and caudal. Its length is 9 inches, but the fish must originally have been a little longer, as its caudal extremity is imperfect.

The scales are very small, and much disarranged, as is the case in nearly all the fishes occurring in the same bed, but their configuration is apparently the same as in N. Greenockii, and their external ornamentation consists, as in that species, of very fine oblique, thread-like, branching, and anastomising ridges; the median ridgescales, extending from behind the dorsal fin along the upper margin of the caudal body prolongation, are very large and pronounced. The bones of the head are much crushed and broken, so that only a few of them can be made out. The suspensorium is very oblique and the gape extensive, the length of the lower jaw being, in the first mentioned specimen, 1 inch. The external surface of the lower jaw is ornamented with a minute and very close tuberculation; the dental margin of the maxilla is also tuberculated, but the rest of its surface is marked with delicate ridges, which run for the most part parallel with its superior and posterior margins. Large conical teeth occur in both jaws, with a few of the external series of smaller ones. One of the larger teeth in the mandible of specimen No. 2 measures a little over inch in length by in diameter at the base; its form is regularly and acutely conical, with wellmarked smooth enamel-cap at the apex, below which the surface is delicately and beautifully striated, the striæ being more pronounced than is usually the case in N. Greenockii. The opercular bones are not visible, but some of the branchiostegal rays may be detected; their outer surfaces are ornamented by very fine flexuous ridges. The paired fins are moderate in size; indeed they might safely be termed small. The pectoral of No. 1 measures one inch in length; in neither specimen can its number of rays be ascertained, but it is evident that the principal ones were unarticulated for about of their length. The length of the ventral in the first specimen is inch; its rays, probably 25 in number, are delicate, smooth, and with their transverse articulations rather far apart. The dorsal and anal fins resemble each other in form, but the latter is apparently slightly the larger; both are triangular, acuminate,