The great body of phenomena comprised under the term "drift," and the smoothed and scratched surfaces of rock, etc., had been by no means unnoticed by the early students of American geology, but they were attributed to violent and widespread water action, and were spoken of in general as "diluvial" formations. When the agency of ice began to be recognized, it was regarded as that of floating and stranding bergs; and this view for a long time contended with the theory of glacial action, even when the latter had been adopted and advocated by eminent students of the subject. The first allusion to drifting ice as the agent of transportation of bowlders, etc., appears to have been made as early as 1825, by one Peter Dobson, of Connecticut, in a letter to Prof. Benjamin Silliman, of Yale College. Sir Roderick Murchison, who became the great champion of this view, credits Mr. Dobson's letter with giving him the first suggestion of it. Twelve years later, in 1837, T. A. Conrad made the earliest reference to land ice as the cause of our drift phenomena; he does this in very striking words when read in the light of the studies and determinations of later years, although of course imperfectly and vaguely. Meanwhile, however, Agassiz and others had been working among the glaciers of the Alps, and their views as to a great period of former extension, in Europe and the British Isles, were finding some acceptance abroad. In this country, Prof. Edward Hitchcock, in his address as retiring president of the Association of American Geologists, in 1841, gave a broad and careful review of the drift phenomena in eastern North America, and referred to the work of Agassiz, Buckland, and Lyell with great interest, as having given him "a new geological sense" in observing these phenomena, and said, with prophetic foresight, "Henceforth, glacial action must form an important chapter in geology."