In the case of the river Somme we have a valley in some places a mile or more in width, and about two hundred feet in depth. This has mostly been excavated in chalk rock. Taking our present large rivers as a basis, it would require from one to two hundred thousand years for the Somme to perform this work.36 It will not do, however, to take the present action of our rivers as a guide, since we have every reason to suppose this work went forward much more rapidly in past times. But we can not escape the conclusion that it demands a very long time indeed to explain it. The valley has remained in its present shape long enough to admit the formation of great beds of peat in some portions. Peat is formed by the decomposition of vegetable growth. Its growth is in all cases slow, depending entirely upon local circumstances. European scholars who have made peat formation a special study assure us that to form such immense beds as occur near Abbeville, several thousand years are required, even under the most favorable conditions. Yet we would be scarcely willing to rest such important conclusions as the foregoing on the researches of one individual, or in one locality. As already stated, DePerthes made his discoveries public in 1847. Yet they were so opposed to all that had been believed previously, that but few took the pains to investigate for themselves. In 1853, Dr. Rigollot, of Amiens, who had been skeptical as to DePerthes, commenced to look for himself in the gravel beds at St. Acheul, about nine miles below Abbeville. As might be expected, he was soon convinced. Paleolithic Flint, England.