GYPSUM OR SECOND FRESHWATER FORMATION. 323

remarkable at Champigny. The compact limestone, thus penetrated with silex, affords by burning, lime of very good quality. We may consider the siliceous limestone, as terminating on one hand the marine formation, and forming on the other, the lower portion of the middle freshwater beds. Hence this rock includes sometimes, in its lower ranges, sea-shells analogous to those of the coarse limestone, mingled with freshwater shells, thus constituting the passing link between the two formations; and the freshwater shells of its upper courses, are the same as those of the middle freshwater limestone. The compact and hard, but easily broken, yellowish limestone, called clicart by the workmen, should be referred to the same locality. In this formation, one kind of the stones well known under the name of mill or buhr stones occurs. These seem to be the siliceous skeleton, of a siliceous limestone. They must not be confounded, however, with the millstones to be described under the eighth head.

5, 6. Gypsum, the second freshwater formation, and marine marls.-These strata afford one of the clearest examples of what is to be understood by a formation in Geology; differing widely from each other in their chemical nature, yet evidently deposited at the same time. This formation which is called gypseous, is not composed solely of gypsum, for it consists of alternate beds of gypsum, argillaceous marl, and limestone, in an order of superposition always the same, over the great gypseous field studied by MM. Cuvier and Brogniart, from Meaux to Triel and Grisy, through

a length of more than 20 leagues. Some of the beds are wanting in a few districts, but the remain. ing ones are always in the same respective positions.

The gypseous course is placed immediately above the marine limestone, a superposition of which it is not possible to doubt. The hills or mounds of gypsum have a peculiar aspect by which they may be recognised at a distance. As they always lie over the limestone, they form as it were second hills of a distinctly oblong or conical shape, situated on the more extended lower elevations.

Both on Montmartre, and the hills which seem to rise in its suite, two formations of gypsum may be observed. The lower is composed of alternate beds of little thickness, consisting of gypsum often crystalline, solid lime-marls, and clay-marls distinctly foliated. It is in the former, that the large crystals of lenticular yellowish gypsum occur, and in the latter the menilite silex. The lower portions of this mass would seem to have been deposited, at one time naked, on the sea-shell calcareous sand, in which case they include marine shells, and at another, on a bottom of white marl, containing a great many freshwater shells, which had previously covered the marine stratum. This second circumstance seems proved by two observations made, the one at Belleville by M. Hericart de Thury, and the other by M. Cuvier under the street of Rochechouart. In digging wells at these two places, the last beds of the lower mass were perforated, and there was found in the under portions of this mass, a great bed of the white freshwater marl above mentioned.

MARVELLOUS SKELETONS IN THE GYPSUM BEDS. 325

Below this bed, occur either the first courses of the marine limestone, or the calcareo-siliceous lacustrine formation.

The superficial mass which the workmen call the first, is in every respect the most remarkable, and important. It is besides much larger than the others; possessing in certain places a thickness of 65 feet, and is uniform throughout, with the exception of a few marly beds. In some places, as at Dammartin, near Montmorency, it lies almost immediately below the vegetable mould.

The lower beds of gypsum of this first mass, include silex, which seems to have flowed into the gypseous matter, and also to be penetrated with it. The intermediate beds naturally split into thick prisms, with several faces, which are called hauts piliers, (high pillars); and the uppermost ranges called chiens are penetrated with marl. They alternate with marl strata and are not very massive. These are commonly five in number, and spread over a great extent.

But these facts are not the most important. The fossils which this mass includes, and those which its marls cover, suggest observations of far higher interest.

It is in this first mass that the quarriers find almost every day, skeletons and scattered bones of a multitude of unknown quadrupeds, along with the bones of birds, crocodiles, the trionyx tortoise, land and freshwater tortoises, and several kinds of fish, most of which belong to freshwater genera.

But what is no less remarkable, and leads to the same conclusion, is that some freshwater shells are

also found there. They are doubtless of very rare occurrence; but a single one would suffice, when unaccompanied with sea shells, to demonstrate the truth of the opinion which Lamanon and some other naturalists broached long ago, that the gypsums of Montmartre, and of the other hills of the Paris basin, had been crystallized in freshwater lakes. We shall presently adduce new facts confirmatory of this judgment.

The superior gypsum is perfectly characterised by the presence of the skeletons and bones of mammifera, which enable us to recognise it, when it occurs insulated. They have never been observed in the inferior masses.

Above the gypsum, are placed thick beds of marl, sometimes calcareous, and at others argillaceous. In the lower courses of these beds, and in a white and friable lime marl, there have been met with at different times, palm-tree trunks petrified with silex. They were recumbent in posture, and of considerable dimensions. In the same system of beds, there occur in almost all the quarries of the Chaumond mound, and even in the quarries to the east of Montmartre, shells of the genera lymnea and planorbis, hardly different from those which now live in our marshes. These fossils prove that the marls are a freshwater formation, like the gypsums placed below them.

In ascending, we encounter a bed of foliated yellowish marl about a yard thick. This forms the limit of the freshwater formation. All the shells that have been observed higher up, are marine.

Immediately over this marl, there is a large and

MARL BEDS WITH BIVALVE SEA-SHELLS. 327

constant bed of greenish clay marl, which may be recognised at a distance by its thickness, its colour, and its continuity. It serves as a guide for finding the cytherea shells, which lie just below it. The four or five beds of marl that succeed the green one, are of little thickness, and like it contain no fossils. But these beds are immediately surmounted by one of yellow clay marl replete with sea shells, of species belonging to the genera cerites, trochus, mactres, venus, cardium, &c. Fragments of the palate of a ray, and portions of the spines of the tail of a ray, were also found in this locality.

Almost all the marl beds which succeed the last, present fossil sea-shells, but only of the bivalve class, and the latter strata, or those immediately below the argillaceous sand, include two distinct beds of oysters. The first and lowest is composed of very thick large oysters; some of them nearly 4 inches long. A layer of whitish marl without shells occurs next, then a second very thick stratum of oysters subdivided into several seams. The oysters are brown, and much smaller, as well as thinner than the preceding. The above oyster beds are almost uniformly present; nor have they been twice missed in the numerous hills examined by MM. Cuvier and Brogniart, so that it seems to be nearly certain that they lived in the place where they now lie; for they are attached to each other, as in the sea, the most of them are quite entire, and when carefully extracted, often retain both their valves. Lastly, M. Defrance has found near Roquencourt, as high as the marine gypseous marls, rounded pieces of the shell marl limestone, pierced with pholades (stone