The second fact, equally well established, and which, though opposed to the opinion of the ancients, can no longer leave any doubt on the mind, is, that rivers, with a few rare exceptions,—such as the Loire for instance,—only carry to the sea an extremely thin mud, destined to be lost in the mass of matter held in suspension by the latter; that the sands of rivers do not in general reach the sea, and that the muddy or sandy deposits observed in tidal rivers, are entirely owing to the matters brought by the tide. This discovery has been arrived at as follows. In making the analysis of the alluvial lands forming the Bay of St. Michael, it was found that the principal substances of their formation are silex and the carbonate of lime; that the nearer the sea is approached, the more the proportion of silex increases; the more it is receded from, the more considerable the proportion of carbonate of lime becomes. Now if the basins of the three rivers which discharge themselves into this bay, the Sée, the Selime and the Couësnon, be examined, they will be found entirely destitute of calcareous substances. It is the same with the coasts of the channel and of Brittany. It cannot, therefore, be either from these rivers or from the coasts that the enormous proportion of silex proceeds which has just been described. If samples are examined with a magnifying glass, commencing with those nearest the sea, and afterwards proceeding farther into the bay, in the first, fragments of shells are perceived quite distinguishable, then these fragments are reduced and become so impalpable, that the best glass will no longer enable us to distinguish the form in the most calcareous portions.