direction, with the mollusks, but he still had nearly all to learn, or, to speak more accurately, nearly all to create, in that uninvestigated world in which Linnæus had failed to introduce the methodical arrangement which he had been so successful in introducing among the higher animals. After devoting a year to preliminary studies, Lamarck began his lectures in the Museum in the spring of 1794; he immediately instituted the great division of animals into vertebrates and invertebrates, which has become fixed in science. Adhering to the Linnæan division of the vertebrates into mammalia, birds, reptiles, and fishes, he divided the invertebrates into mollusks, insects, worms, echinoderms, and polyps. In 1799 he separated the order of crustaceans from the insects with which it had been confounded; in 1800 he separated the arachnids from the insects; in 1802 he set off the annelids as a subdivision of the worms, and the radiates as separable from the polyps. Time has confirmed the justice of his division, which depends in every respect upon the organization of the animals. This is the rational method, incorporated in science by Cuvier, Lamarck, and Geoffroy Saint-Hilaire.
As our sketch has so far dealt only with Lamarck's achievements in natural history, we pass with a simple mention a few works in which he treated of physics and chemistry; mistakes of a good intention, which attempted to establish truths that rest exclusively on experiment, by reasoning alone, or to resuscitate old theories like that of phlogiston. These efforts did not even receive the honor of a contradiction; they did not deserve it; and they should serve as a warning to all those who would write upon any science without being acquainted with it, and without having had practical experience in it.
The generalizations of Lamarck in geology and meteorology, sciences which at the time he wrote had hardly come into existence, were mistaken in another sense. They were premature. Every science must begin with the knowledge of facts and phenomena. When these are numerous enough, a partial generalization is possible; as they increase, the basis grows broader; but systems which can justly claim to be absolute and definitive can never be, for they presuppose that all the phenomena and facts are known, a condition which will be impossible as long as man lives. In the beginning of this century geology did not exist, and little was known of the matters of which it treats; but systems were created that included the whole earth. Lamarck elaborated his system in 1802; and twenty-three years afterward the clear mind of Cuvier succumbed to the prevailing tendency, and he published his treatise on the revolutions of the globe. It was Lamarck's merit that he perceived that there were no revolutions in geology, and that the slow manifestations of force through hundreds of thousands of years far better explained the wonderful changes of which our planet has been the scene than violent disturbance could do. "To nature," he said, "time is nothing: it is no obstacle. Nature