in the previous two sections and those recorded in § 80, to see that in the present case a double anhydrous carbonate is actually generated. While in the previous preparations with sulphate of magnesia and carbonate of lime, or with the more hydrated double carbonate, the separation of the two carbonates by dilute acid is nearly complete, a large amount of carbonate of magnesia is in this case dissolved with the first portion, and the residue is, by the farther action of the dilute acetic acid, shown to be a mixture of dolomite with magnesite.

108. But the most favorable conditions for the artificial production of dolomite, so far as yet observed, are attained with an intimate mixture of the two carbonates in the amorphous state, as precipitated by a slight excess of carbonáte of soda from the solution of equivalent proportions of the chlorids of calcium and magnesium. (§82). To effect the union of the two carbonates the heat should be very gradually raised to 120°-130°, and there maintained for an hour or two. The following are results obtained from preparations thus obtained and submitted to fractional analysis by means of acetic acid of three per cent.

A precipitate of the two carbonates was made as above, pressed in linen, and the unwashed pasty mass gradually heated, thirty minutes being taken to raise the temperature from 100° to 140° C. Of the washed and dried matter the first thirty-four per cent contained 566, and the last ten per cent 53.0, of carbonate of lime. Another and partially washed precipitate of equivalents of the two carbonates, which had been dried at the ordinary temperature, and again moistened with water and heated to 170° C., was treated with successive portions of acetic acid with similar results to the preceding. The last residue of twenty-one per cent consisted of carbonate of lime 52.7, carbonate of magnesia 47.3.

It is unnecessary to multiply the descriptions of results of this kind obtained from five or six different preparations, and all showing that under the influence of heat the pasty mixture of the two carbonates yields an anhydrous, sparingly soluble compound having the chemical character and composition of dolomite, which requires carbonate of lime 54.35, carbonate of magnesia 45.65.

§ 109. In another experiment a mixture containing more than an equivalent of magnesian carbonate was heated as above described, and the portion dissolved by the first action of the acid contained 48.6 per cent of carbonate of magnesia, while the second portion dissolved had only 470 per cent, and the residue was pure magnesite. The excess of magnesia in the first fraction over the second would seem to be due, as in § 105, to a partial decomposition of the excess of hydrated magnesian carbonate in the mixture.

§ 110. Carbonic-acid 'water may be employed instead of acetic acid as a solvent for the analysis of artificial dolomite ($79). In a preparation of dolomite made in the way just described, and containing an excess of carbonate of magnesia, (520 per cent), the action of 500 c. c. of water saturated with carbonic acid during two and a half hours, removed from one gram 0453 gr. containing only 48.5 per cent of magnesian carbonate. The residue, from which the more finely divided portions had thus been removed, was very slowly attacked by a solution of carbonic acid, a second portion of 500 c. c. of which, after four hours, took up 0.145, and a third portion, after eighteen hours more, 0.162 gr. of the two carbonates, in both cases consisting of carbonate of lime 53 0, carbonate of magnesia 47.0.

$111. In concluding this part of the subject it is to be remarked that two things in the history of dolomite may be regarded as established: first, its origin in nature by direct sedimentation, and not by the alteration of non-magnesian limestones; and second, its artificial production by the direct union of mixtures of the carbonates of lime and magnesia at temperatures above 120° C. The question next arises whether all dolomite strata have been exposed to such a temperature, or whether there are yet unknown conditions under which the double carbonate can be found at lower temperatures.

The magnesian limestone from the elevated coral island of Matea, described by Dana (this Jour., [2], xiv, 82), is, according to the analysis of Silliman, and my own subsequent examination and analysis (Ibid., [2], xix, 429), a true dolomite with a slight excess of carbonate of lime, and is regarded by Dana as of recent origin, and as derived, in some way, from the alteration of coral mud. If this origin be established beyond a doubt, it is to be remarked that the separation of carbonate of magnesia from sea-water requires peculiar conditions, which evidently are rarely fulfilled in the case of these coral deposits; and its production being conceded, the volcanic agencies so active in these regions may have very well furnished the heat requisite to form dolomite before the elevation of the island.

$112. Apart from the formation of stratified sedimentary dolomite, we have also to keep in mind the frequent occurrence of this double carbonate as a mineral of secondary deposition, lining drusy cavities, filling veins, and even the moulds of fossil shells (§ 52, 53). The conditions of its deposition from natural waters are probably not unlike those of the quartz, fluor, and heavy-spar, with which, in its form of bitter-spar, it is often associated, and as subjects for farther investigation, may yet throw more light on the agencies which have effected the union and crystallization of the two carbonates in sedimentary deposits. Montreal, Jan. 1866.

ART. VIII.-Remarks on the new division of the Eocene, or Shell Bluff Group, proposed by Mr. Conrad; by EUG. W. HILGARD, Ph.D., State Geologist of Mississippi.

IN a brief paper published in the January number of this Journal, Mr. Conrad proposes to distinguish, as a separate group of the American Eocene, a series of deposits but feebly represented at Vicksburg by a five-foot stratum of dark lignitic clay and sand, differing in its paleontological characters from both the Vicksburg and Jackson group. Mr. Conrad considers it to be especially characterized by the occurrence of Ostrea Georgiana, and defines it as underlying the "Orbitolite limestone of the Jackson Group." He also mentions, in the section of the Vicksburg Bluff, the Orbitolite limestone, as a representative of the Jackson group.

The latter supposition is manifestly an oversight on the part of my honored friend. That the group of fossils described by him, and figured in Prof. Wailes's Report, as Jackson fossils, do not occur at Vicksburg, I need not recall to his mind; but he has overlooked the fact that the Orbitoides Mantelli, throughout the state of Mississippi, at least, is entirely absent from the Jackson Group, the Orbitoides limestone being invariably accompanied by Pecten Poulsoni, Arca Mississippiensis, Ostrea Vicksburgensis, and other leading Vicksburg fossils.

Of Ostrea Georgiana I have unfortunately never seen an authentic specimen or description; but from the facts stated by Mr. Conrad, and his comparing it to P. longirostris Lamk., I unhesitatingly identify it with specimens from Vicksburg, labeled "P. gigantea" by Prof. Wailes. Upon the authority of the latter observer, Mr. Conrad mentions the occurrence of O. Georgiana at Jackson. There is, indeed, some resemblance between the lower valve of the oyster so abundant at Jackson (which, together with the bones of the Zeuglodon, characterizes the "shell prairies" of central Mississippi, as stated p. 128 of my Report, and some specimens of O. Georgiana; but in their general character, when seen in series, they differ widely, the Jackson oyster having distinctly the habitus of a Gryphia, and oftentimes resembling closely G. convexa of the Rotten limestone. It is one of the leading fossils of what I have most unequivocally recognized as the upper member of the Jackson group; it occurs at Jackson itself, on the hill-tops, associated with Zeuglodon bones, Umbrella planulata, Cypræa fenestralis, Morio Petersoni, Conus tortilis, and others, in stratum No. 7 of section 27, page 131 of my Report. The Jackson fossils described by Mr. Conrad are derived from Nos. 4 and 5 of that section.

In the numerous localities where I have studied the beds of the Jackson group, I have never found a single Orbitoides associated with them. The constant concomitant of the latter fossil, the Pecten Poulsoni, also is absent from the Jackson strata, being replaced by P. nuperus, Con.

But if the Orbitoides limestone is no member of the Jackson, but on the contrary, a characteristic one of the Vicksburg group, then it is clear that the strata of the "Shell Bluff group" at Vicksburg lie above, and not below the Jackson strata. For it cannot be supposed that the latter, which occupy so extensive. an area above Vicksburg (see the map accompanying my Report,) should suddenly come to an end, and leave no trace of a representative between the Shell Bluff and the Vicksburg groups did it belong there.

There is only one other locality in the state, as far as known, where O. Georgiana (i. e. the large oyster occurring at Vicksburg) is found, viz: in Jasper county, Miss., where it was collected by Prof. W. D. Moore, late of the University of Miss. It there occurs again in the same outcrop with Pecten Poulsoni, Orbitoides, and a Schizaster, which is also a leading Vicksburg fossil; this locality being, likewise, considerably south of the shell prairies of the Jackson group.

As there is nothing to justify the assumption of a sudden termination of the strata of the latter group, which, on the contrary, may be seen disappearing under those forming the transition to the Vicksburg strata, with remarkable regularity, along the course of both Pearl and Chickasawhay rivers, (see p. 135 of my Report), the conclusion is inevitable that the Jackson group is older than the Shell Bluff group as defined by Conrad.

That there may be a considerable difference in the geological horizons of the Jackson and Vicksburg groups proper, sufficient to admit of the existence of a fauna deserving to be formed. into a distinct group, is proved, not only by the paucity of coincident species, (see list, ibid, p. 132), but no less by the considerable thickness of the intervening strata in eastern Mississippi, on the Chickasawhay river, which near Red Bluff Station (ibid, p. 135,) amounts to over one hundred feet.

Here, as at Vicksburg, we have, underlying the Orbitoides, marls and limestones, a stratum of inconsiderable thickness, but literally teeming with shells, which are a strange mixture of the faunas of Jackson and Vicksburg, with numerous peculiar species (see list, ibid, p. 136). Here also, we have a Madrepora, distinct from, but closely allied to, the species found in the "Georgiana bed" at Vicksburg; where in its turn we find an extraordinary number of valves of Meretrix sobrina, a rara avis in the Vicksburg strata proper, but abundant in the Jackson group. Busycon undulatum, also, is a Jackson form, if not

species. The number of species thus far found in this bed is, however, too small to allow us to expect numerous coincidences either way. Ostrea Georgiana does not occur at Red Bluff, the greater part of whose strata, immediately superimposed upon the Zeuglodon beds, are extremely poor in fossils.

Of course, these data are insufficient as yet to parallelize Mr. Conrad's Shell Bluff and my Red Bluff group. But their relative positions seem to be at least analogous, and I sincerely hope Mr. Conrad will, before many months, give both the collections of the Mississippi Survey, and the localities themselves, the benefit of his personal inspection.

I should add, in conclusion, that the Jackson and Vicksburg groups are by no means always thus separated by intervening beds. In one locality, at least, at the extreme southern edge of the marine Tertiary, I have seen the white Orbitoides marl directly superimposed upon a bed of blue marl containing Monoceros vetustus and Morio Petersoni, with a Rostellaria thus far called R. velata, but which is certainly a species distinct from the Claiborne fossil of that name.

ART. IX.-Preliminary Notice of certain beds of Fish-remains, in the Hamilton group of Western New York; by FRANK H. BRADLEY.

ACCOMPANYING the Moscow black shale of the upper part of the Hamliton group, at Geneseo and Moscow, Livingston county, and East Bethany, Orleans county, N. Y., are certain thin lenticular masses of impure pyrites, which contain large quantities of the teeth, fin-spines and bony-scales, of fishes, and numerous Mollusca.

The layers composing these beds are very variable in thickness and in composition, some being quite solid and composed almost entirely of pyrites; others, thin and fragile, and interlaminated with layers of black shale. The latter portions commonly contain the bones, while the more solid portions yield shells most numerously.

It would seem that the sulphur of the pyrites must have come from the decomposed fish, and that the beds correspond to the deposits of fish-remains reported by dredgers in certain seas, while the surrounding bottom yields not a fragment.

Information concerning the situation of these localities was given by Mr. H. A. Green in the January number of this Journal.

So far as I have been able to ascertain, they had not been explored by any one previous to my visit in July, 1864, at which