This plain is on the east side of the Rhone, between Salon and Aries: it is of a triangular form, about twenty square leagues in extent, and is covered almost entirely with quartzy gravel. This immense collection of gravel has been supposed by some to have been brought down by the Durance from the Alps of Dauphine; by others it has been ascribed to the Rhone; and by many to the sea, as being a work too great for any river. The explanation mentioned above, § 105, namely, that the loose gravel on the plain arises from the decomposition of a great stratum of pudding-stone, which is the basis of the whole, is the opinion of Saussure, and is founded on his own observations.[167] [Footnote 167: See Voyages aux Alpes, tom. iii. § 1592 et 1597. See also on this subject a Memoir by Lamanon, Journal de Physique, tom. xxii. p. 477; and another by M. De Servieres, _ibid._ p. 270.] 335. The theories that have been contrived for explaining the phenomena of the plain of Crau, afford an instance of the necessity of generalizing our observations before we can explain a particular appearance: in other words, they prove the truth of Lord Bacon's maxim, That the explanation of a phenomenon should not be sought for from the study of that phenomenon alone, but from the comparison of it with others. One of the theories of this plain is, that the breccia, which is the base of it, is formed from the consolidation of the loose gravel of the plain, by water percolating through it, and carrying some cementing substance along with it, or some _lapidific juice_, as it is called. And indeed, whether the gravel is formed from the breccia, or the breccia from the gravel, is a question which probably could never be resolved by the mere examination of the plain itself. But the question is very soon decided, when we compare what is observed here with other appearances in the natural history of the earth's surface, and consider how much more frequent the decomposition of solids is, than their reconsolidation, in any place above the level of the sea.