The gypsum and salt deposits of the Salina show that toward the close of the Silurian period a slight oscillation brought the sea floor nearer to the surface, and at the north cut off extensive tracts from the interior sea. In these wide lagoons, which now and then regained access to the open sea and obtained new supplies of salt water, beds of salt and gypsum were deposited as the briny waters became concentrated by evaporation under a desert climate. Along with these beds there were also laid shales and impure limestones. In New York the "salt pans" of the Salina extended over an area one hundred and fifty miles long from east to west and sixty miles wide, and similar salt marshes occurred as far west as Cleveland, Ohio, and Goderich on Lake Huron. At Ithaca, New York, the series is fifteen hundred feet thick, and is buried beneath an equal thickness of later strata. It includes two hundred and fifty feet of solid salt, in several distinct beds, each sealed within the shales of the series. Would you expect to find ancient beds of rock salt inclosed in beds of pervious sandstone? The salt beds of the Salina are of great value. They are reached by well borings, and their brines are evaporated by solar heat and by boiling. The rock salt is also mined from deep shafts. Similar deposits of salt, formed under like conditions, occur in the rocks of later systems down to the present. The salt beds of Texas are Permian, those of Kansas are Permian, and those of Louisiana are Tertiary. =The Mississippi valley.= The heavy near-shore formations of the Silurian in the Appalachian region thin out toward the west. The Medina and the Clinton sandstones are not found west of Ohio, where the first passes into a shale and the second into a limestone. The Niagara limestone, however, spreads from the Hudson River to beyond the Mississippi, a distance of more than a thousand miles. During the Silurian period the Mississippi valley region was covered with a quiet, shallow, limestone-making sea, which received little waste from the low lands which bordered it. The probable distribution of land and sea in eastern North America and western Europe is shown in Figure 287. The fauna of the interior region and of eastern Canada are closely allied with that of western Europe, and several species are identical. We can hardly account for this except by a shallow-water connection between the two ancient epicontinental seas. It was perhaps along the coastal shelves of a northern land connecting America and Europe by way of Greenland and Iceland that the migration took place, so that the same species came to live in Iowa and in Sweden. [Illustration: Fig. 287. Hypothetical Map of Parts of North America and Europe in Silurian Time. Shaded areas, probably seas; broken lines, approximate shore lines]