It seems to be curiously difficult to obtain fair treatment for the theory of a continuously cooling earth and its consequence, the thermal contraction theory of mountain formation. After Davison and Darwin had calculated the available compression of the earth's crust on Kelvin's theory of the cooling of the earth, and obtained a result that was at any rate of an order of magnitude consistent with the observations then available, Osmond Fisher made a revised comparison, which amounted to supposing the mountains formed by compression all powdered and spread out uniformly over the whole surface of the earth; the thickness of the resulting layer was naturally very much less than the actual height of known mountains. So far as I have been able to trace, nobody protested against the assumption that the thickness of the material after it had been redistributed in this way was directly comparable with the height of the mountains as actually distributed, and Osmond Fisher's view seems to have been universally quoted as the last word on the subject. Even long after my rediscussion of the whole question in 1916, making allowances for the additional data provided by radioactivity, had again led to a result agreeing with observation within the uncertainty of the data, it was almost completely neglected in the literature of the subject, while Osmond Fisher's work retained undisputed pre-eminence.