Jean André Deluc in 1822 published a paper discussing the theories of MM. de Cossigny, Prévost and Pictet. Deluc had never visited a glacière himself, but he explains clearly the impossibility of Professor Pictet's cold current theory, on the simple ground that Professor Pictet himself did not find any cold currents. He takes up Professor Prévost's theories warmly; using also the manuscript notes of Mons. Colladon who had visited the Grand Cave de Montarquis. Deluc says: "that the winter's cold penetrates into these caves, freezes the water which collects there and that the ice thus formed has not the time to melt during the following summer." He says further: "It seems that in the three glacières with which we have been occupied there is a flat or rather hollow bottom, where the waters can form a more or less deep pond, and whence they therefore cannot flow away; it is there they flow in winter; and as these are shut in places where the air cannot circulate, the heats of summer can only penetrate very feebly. The ice once formed in such cavities, only melts slowly; for one knows that ice in melting, absorbs 60° of heat; and where find this heat in an air always very cold and nearly still? During a great cold, the ice forms with great promptness, while it melts with much slowness, even when the temperature of the air is several degrees above zero; what must then not be this slowness when the temperature of the interior air only rises in summer one degree above freezing point. It would need several summers to melt this ice if it did not reform each winter." C. A. Lee, in 1825, wrote that the ice in the Wolfshollow near Salisbury was a winter formation. G. Poulett Scrope, in 1826, accepted as the truth the statement that the cave of Roth was filled with ice in summer, but that it was warm during the winter. In 1827, he explained the presence of ice at Pontgibaud as follows: "The water is apparently frozen by means of the powerful evaporation produced by a current of very dry air issuing from some long fissures or arched galleries which communicate with the cave, and owing its dryness to the absorbent qualities of the lava through which it passes." F. Reich, in 1834, thought that there were two possible causes which might produce subterranean ice: 1, the difference in specific gravity between warm and cold air; 2, evaporation. He thought the cold air a sufficient cause in most caves, but he considered that evaporation also played a part not infrequently. Professor Silliman, in 1839, gave the first hint, in the negative, about compressed air as a cause for subterranean ice. He said about Owego that if one could suppose that compressed gases or a compressed atmosphere were escaping from the water or near it, this would indicate a source of cold, but that as there is no indication of this in the water, the explanation is unavailable.