It has long been known that wherever the calcareous strata are sufficiently hard and compact to support a roof, caves are to be found in greater or less abundance. Those of Devonshire occur in the Devonian limestone; those of Somerset, Nottinghamshire, Yorkshire, Derbyshire, and Northumberland, as well as of Belgium and Westphalia, in that of the carboniferous age. In France also, those of Maine and Anjou, and most of those of the Pyrenees and in the department of Aude, are hollowed in carboniferous limestone, as well as the greater part of those in North America, in Virginia, and Kentucky. The cave of Kirkdale in Yorkshire, and most of those in Franconia and in Bavaria penetrate Jurassic limestones, which have received the name of Hohlenkalkstein from the abundance of caverns which they contain. They are developed on a large scale in the Swiss and French Jura, and in some cases afford passage to powerful streams, and in others are more or less filled with ice, thus constituting the singular “glacières” that have been so ably explored by the Rev. G. F. Browne.[22] The compact Neocomian and Cretaceous limestones contain most of the caverns of Périgord, Quercy, and Angoumois, and some of those in Provence and Languedoc, those of Northern Italy, Sicily, Greece, Dalmatia, Carniola, and Turkey in Europe, of Asia Minor and Palestine. The tertiary limestones, writes M. Desnoyers,[23] offer sometimes, but very rarely, caves that have become celebrated for the bones which they contain, such as those of Lunel-Viel, near Montpelier, those of Pondres and Souvignargues, near Sommières (Gard), and of Saint Macaire (Gironde). The same may also be said of the calcaire grossier of the basin of Paris. Certain rocks composed of gypsum also contain caverns of the same sort as those in the limestones. In Thuringia, for example, near Eisleben, they occur in the saliferous and gypseous strata of the zechstein, and are connected with large gulfs and cirques on the surface, which are sometimes filled with water. In the neighbourhood of Paris, and especially at Montmorency, they contain numerous bones of the extinct mammalia. M. Desnoyers points out their identity, in all essentials, with those in calcareous strata, and infers that they have been produced in the same way. Some of them may have been formed by the removal of the salt, which is very frequently interbedded with the gypsum, by the passage of water. In Cheshire the pumping of the brine from the saliferous and gypseous strata produces subterranean hollows, which sometimes fall in and eventually cause depressions on the surface, such as those which are now destroying the town of Northwich, and causing the neighbouring tidal estuary to extend over what was formerly meadow land. This explanation, however, will not apply to those in the neighbourhood of Paris, because there is no trace of their ever having contained salt. _The Relation of Caves to Pot-holes, “Cirques,” and Ravines._