While, then, the products of animals such as corals and molluscs are limestones, those of vegetables are humus and clay; and all of these deposits losing their less fixed principles pass into a silicious condition, and end by being reduced to quartz, which is the earthy element in its purest form. The salts, pyrites, and metals only differ from other minerals by the different circumstances under which they were accumulated, in their different proportions, and in their much greater amount of carbonic or acidific fire. Regarding granite, which, he says, naturalists very erroneously consider as _primitive_, he begins by observing that it is only by conjecture that we should designate as primitive any matter whatever. He recognizes the fact that granite forms the highest mountains, which are generally arranged in more or less regular chains. But he strangely assumes that the constituents of granite, _i.e._, felspar, quartz, and mica, did not exist before vegetables, and that these minerals and their aggregation into granite were the result of slow deposition in the ocean.[80] He goes so far as to assert that the porphyritic rocks were not thus formed in the sea, but that they are the result of deposits carried down by streams, especially torrents flowing down from mountains. Gneiss, he thinks, resulted from the detritus of granitic rocks, by means of an inappreciable cement, and formed in a way analogous to that of the porphyries. Then he attacks the notion of Leibnitz of a liquid globe, in which all mineral substances were precipitated tumultuously, replacing this idea by his chemical notion of the origin of the crystalline and volcanic rocks. He is on firmer ground in explaining the origin of chalk and clay, for the rocks of the region about Paris, with which he was familiar, are sedimentary and largely of organic origin. In the "Addition" (pp. 173-188) following the fourth chapter Lamarck states that, allowing for the variations in the intensity of the cause of elevation of the land as the result of the accumulations of organic matter, he thinks he can, without great error, consider the mean rate as 324 mm. (1 foot) a century. As a concrete example it has been observed, he says, that one river valley has risen a foot higher in the space of eleven years. Passing by his speculations on the displacement of the poles of the earth, and on the elevations of the equatorial regions, which will dispense with the necessity of considering the earth as originally in a liquid condition, he allows that "the terrestrial globe is not at all a body entirely and truly solid, but that it is a combination (_réunion_) of bodies more or less solid, displaceable in their mass or in their separate parts, and among which there is a great number which undergo continual changes in condition."