invasion. These lines of boulders are to be seen in many places, registering accurately, not the work of the preceding year, but the greatest effort of any previous year.

The circumstances of some deep-lying boulders may be such that they are rarely embraced, acted on, or moved, and such may long, by fits, continue to be erratic, though finally to join the general shore parade.

The force of these moving fields is very great, even when the decomposing process is much advanced. I have seen a timber wharf, which was about thirty feet square, ten or fourteen feet high, and filled solidly with earth and stones, shoved along the bottom about thirty feet, by a single continuous push of a great field of ice just ready to be resolved into its prismatic elements. The motion was very slow, only to be seen, indeed, by close observation, while the ice was broken at the edge of contact into innumerable fragments, piling themselves, with a tinkling sound, high upon the wharf and following ice.

A simple and effectual guard against this danger to wharf or pier has been found to be, the giving to the exposed face a certain talus (about one of base to two of height, I think), which turns the ice upwards to the top of the structure, where its fragments accumulate, sometimes to a considerable height. This easy diversion of so great a force is due, of course, to the peculiar crystalline structure of the ice, the degree to which it has been decomposed, and the consequent brittleness against a transverse strain. Should there be an unfrozen margin to permit this motion of large fields of ice, before the solution of continuity in the crystalline arrangement, nothing but the solid earth could stand before it.

These remarks have extended further than I intended, and I fear much beyond what was required by the state of knowledge on the subject. But I venture, nevertheless, in reference to the first portion of these remarks, one further observation—namely, that nature seems to have especially provided, in the structure of these wintry coverings of water surfaces, for their prompt removal, when their existence would retard the advancing year.

ART. XL.-On some Reactions of the Salts of Lime and Magnesia, and on the Formation of Gypsums and Magnesian Rocks; by T. STERRY HUNT, F.R.S., of the Geol. Survey of Canada. (Continued from this vol., p. 187.)

IV.

Facts in the history of Gypsums, Dolomites, Magnesites and Lime

stones.

43. The gypsums found in nature may be divided into two classes, those directly deposited from water, and those produced by the alteration of beds of limestone. To the latter division belong the gypsums found in the vicinity of solfataras, where, as Dumas has shown, the slow oxydation of moist sulphuretted hydrogen gives rise to sulphuric acid, which transforms beds of carbonate into hydrated sulphate of lime. We must equally refer to the same class those gypsums which are formed among calcareous rocks by the action of waters containing free sulphuric acid. Such a process I have long since described in Western Canada, where numerous springs containing besides sulphates of lime, magnesia, oxyd of iron, alumina, and sulphuretted hydrogen, three or four thousandths of free sulphuric acid, rise through Upper Silurian strata, in the calcareous portions of which they sometimes give rise to masses of gypsum.

Bischof (Chem. Geology, i, 418), who does not appear to have seen my analyses of these acid waters, rejects my view of the epigenic origin of these masses of gypsum, although it will be apparent to every one who examines the facts, that the action of such waters upon calcareous strata must give rise to sulphate of lime. I do not however confound these recently formed masses of sulphate of lime with the older gypsums, which associated with dolomites, sea-salt and sulphur, are abundant in the Saliferous or Onondaga salt group of the same region.-(Am. Jour. Sci., [2], vii, 175; Report Geol. Survey, 1848, 150; Comptes Rendus de l'Acad., 1855, xl, 1348.)

These acid waters which make their appearance in an almost undisturbed region, I conceive to have their origin in deeply buried strata, where gypsum or other sulphates may be undergoing decomposition by the action of water and silica at an elevated temperature, a process analogous to that which gives rise to exhalations of carbonic acid gas.

44. Waters containing free sulphuric acid or ferric or aluminous sulphate, may by flowing into basins where carbonate of lime is present, give rise to solutions of sulphate of lime, and the evaporation of these, of sea-water or other gypseous solutions must give rise to deposits of sulphate of lime, which will

belong to the first division mentioned above. These modes of formation however do not account for an important fact in the history of most stratified gypsums, which is that of their almost constant association with carbonate of magnesia generally in the form of magnesian limestone. Beds of dolomite are often interstratified with or include beds or masses of gypsum, while dolomite and carbonate of magnesia are sometimes found imbedded in gypsum or anhydrite. For a description of the magnesite which is disseminated in the gypsum of Salzburg, see Dufrénoy, Minéralogie, 2d ed., ii, 424. Small masses of compact and crystalline gypsum, occasionally associated with crystals of calcite and quartz, abound in some of the dolomite beds of the so-called Calciferous sandrock in Canada, and crystallized gypsum and anhydrite, together with sulphates of baryta and strontia, and fluor spar, occur in geodes in the magnesian limestone of Niagara. The anhydrous sulphate of lime not only forms beds by itself but is often met with disseminated in masses, grains or crystals through beds of gypsum, and even interstratified with it, as in the south of France, in the Hartz, Switzerland, and in Nova Scotia, as described by Mr. Dawson. (Acadian Geology, 225.) The conversion of beds of anhydrite into gypsum by the absorption of water, and the attendant phenomena, have been described by Charpentier.

45. Both the hydrous and anhydrous sulphate sometimes form the cement of conglomerates or breccias, which enclose flints, fragments of shale and of limestone, as at Pomarance in Tuscany, (Scarabelli, Bull. Soc. Geol. de France, [2], xi, 346,) and also at Bex, where the cement of the conglomerate is a granular anhydrite (Charpentier, Ibid., [2], xii, 546).

Gypsums moreover often include clay and sand, and sometimes contain a considerable admixture of carbonate of lime, which in those of Aix, according to Coquand, amounts to eight per cent. The gypsums of Montmartre also contain, according to Delesse, besides some clay and sand, and several hundredths of carbonate of lime, not less than three per cent of soluble silica intermixed. Silica in the form of flint or chert is sometimes found in concretions with gypsum; thus in the miocene clays near Bologna in Italy, flints are met with associated with sulphates of lime, of baryta and strontia, together with pyrites and sulphur. Masses of sulphate of strontia are likewise found in clays with the gypsums of Montmartre, and the association of sulphate of strontia with the sulphur, gypsum and rock salt of Sicily is well known. The gypsums of Madrid, which occur in tertiary clays, are according to Casiano de Prado, accompanied by beds of chert and of magnesite (Bull. Soc. Geol. de France, [2], xi, 334).

Besides the rock salt which so often occurs with gypsum, we may here recall its frequent association with the sulphates of soda and magnesia, both of which are found in very many places imbedded in gypsum, or intermingled with rock-salt or with the associated clays. (Bischof, Chem. Geology, ii, 421-431.) Large deposits of both of these sulphates occur with gypsum and rocksalt in Spain; in Nova Scotia also sulphate of soda is found in gypsum with boro-calcite, an association worthy of notice from the occurrence of boracite, both crystallized and massive (stassfurthite) with gypsums in Germany.-(How, Am. Jour. of Science, [2], xxiv, 230.)

46. The gypsums of the class which we are now describing appear in every geological period. To these apparently belong the masses of gypsum and anhydrite, which at Fahlun are associated with dolomite and serpentine in the chloritic bands of the oldest crystalline rocks of Scandinavia, the probable equivalents of the Laurentian system of North America. On this continent the oldest known gypsums are those already mentioned as occurring near the base of the paleozoic series, and in what is called by the geologists of New York the Calciferous sandrock. As we ascend the series gypsum is occasionally met with in the Clinton and Niagara groups, until we reach the Onondaga salt-group in the Upper Silurian rocks of Canada and New York, which contains great deposits of dolomite and gypsum, occasionally accompanied by sulphur. The gypsums, anhydrites, and brine springs of Nova Scotia belong to the Carboniferous series, while the frequent recurrence of gypsum in Europe through all the higher rocks up to the Miocene inclusive, is too well known to require

notice.

47. The so-called primitive gypsums and anhydrites, which in the Alps and Pyrennees occur interstratified with crystalline schists, are now known to belong to altered secondary strata. These gypsums enclose many crystalline minerals, such as talc, mica, epidote, hornblende, dipyre, beryl, quartz, hematite, blende and pyrites. At Saurat in the Pyrennees many of these minerals appear in the vicinity of a mass of granite which penetrates and alters both fossiliferous limestone and gypsum. The latter becomes mingled with and finally passes into limestone. (Coquand, Bull. Soc. Géol. de France, [1], xii, 345.) In Algiers, where gypsum is associated with crystalline limestone, gneiss, amphibolite and serpentine, small crystals of beryl are found disseminated alike through the limestone and the gypsum. Some of the gypsums of the Hartz, according to Frapolli, contain nodules of a silicate of magnesia colored by carbonaceous matter, and having the softness and the chemical composition of steatite.-(Ibid., [2], iv, 832.)

48. The marine origin of the greater number of gypsiferous formations is evident both from the accompanying rock salt and the associated fossils, but certain gypsums (as well as certain dolomites,) have evidently been deposited in fresh-water basins. A gypsum from Asia Minor examined by Ehrenberg contains a great number of fresh-water polygastric infusoria, and beds of gypsum occur in the lacustrine basins of Aix and of Auvergne; the gypsiferous strata of the Paris basin are also regarded as of fresh-water origin.

49. Besides the magnesian limestones of gypsiferous strata great deposits of dolomite occur in the rocks of every geological period. I have long since described the dolomites which form extensive beds, often associated with ophiolites and with crystalline limestones, in the Laurentian system in Canada. Great portions of the paleozoic limestones of North America are magnesian, especially in the valley of the Mississippi,* while deposits of dolomites are found in Europe alike in the Permian, Triassic, Jurassic, and Tertiary strata. Mr. Dana has even described as of recent formation a dolomite from the coral island of Matea, examined by Silliman and myself.—(Am. Journal of Science, [2], xix, 429.)

50. The mechanical conditions of these magnesian limestones vary greatly; they are sometimes made up of crystalline grains of dolomite, which are strongly coherent, or more rarely form a loose sand. Not unfrequently the magnesian limestones are concretionary in their structure, and may be oolitic or botryoidal. The action of the concreting force has sometimes obliterated the marks of stratification. The porous or cavernous structure of many dolomites is also to be remarked.

For the following facts with regard to the dolomites of the paleozoic rocks of the Mississippi valley, I am indebted to Prof. James Hall of Albany. We have there in ascending order:

1st. The so-called Lower Magnesian limestone, which is regarded as the equivalent of the Calciferous Sandrock, and is from 200 to 250 feet in thickness. It is the lead-bearing rock of Missouri, and probably contains the cobalt ores of that region. 2d. The Galena limestone, consisting of about 250 feet of dolomite interposed between the Trenton and the Hudson River groups. It is the lead-bearing rock of Iowa, Wisconsin and Illinois.

3d. The Niagara limestone, also dolomitic, about 250 feet in thickness, and sometimes holding galena and blende.

4th. The Leclaire or Galt limestone, a dolomitic formation interposed between the last and the Onondaga Salt Group. It attains upon the Mississippi a thickness of 500 feet, but thins out to the eastward.

5th. The magnesian limestones of the Onondaga salt group, 100 feet thick. 6th. A dolomitic deposit in the upper part of the Carboniferous series. The formation No. 1, although generally regarded as the equivalent of the Calciferous sandrock, is perhaps the representative of the Chazy limestone, which on Lake Huron is sometimes a pure dolomite, and on the island of Montreal includes thin magnesian beds. The Calciferous sandrock itself, throughout Lower Canada, includes extensive beds of dolomite, and the Hudson River group is characterized by beds of dolomite and of magnesite.