Royal Society. 143 



seen that refrigeration has the effect of precipitating the salts of sea- 

 water ; and further, that it appears to him very probable that the 

 temperature at which water begins to expand by the continued ap- 

 plication of cold is that at which saline and earthy matter begins to 

 be precipitated in solutions of the density of sea-water. 



From the immense depth to which icebergs extend in Davis's 

 Straits, and also from their vast number, the author infers that the 

 temperature of the water will be kept pretty uniformly the same 

 throughout a considerable part of its depth, rarely exceeding + 32°, 

 except at the surface, where the action of the sun comes into ope- 

 ration, in which case the water of greatest density from saline con- 

 tents would always occupy the lowest position. In illustration of 

 his views, he describes experiments on the freezing of sea- water of 

 the density 1*025, in glass tubes ; and from these he infers that, not 

 only does congelation precipitate the saline matter in water, but 

 refrigeration also at temperatures from 40° down to 32°. With re- 

 ference to the influence of the density of the sea- water on currents, 

 he remarks that after the warm season has fairly set in, in the Arctic 

 seas, nothing is more common than to observe the surface-water, in 

 hollowed out lanes or fissures of the land-ice, moving slowly towards 

 the open water at the edge of the fixed ice ; and this seaward mo- 

 tion is altogether independent of tidal motion or oceanic current, 

 depending entirely upon the diminished density of the surface-water. 



In conclusion, the author states that he does not know that we 

 are yet in a position to demonstrate the actual existence of currents 

 into the icy seas, as well as out of them ; but that the necessity for them 

 is obvious. It is not necessary, he remarks, that these currents, as in 

 other parts, should occupy the surface, and probably also the bottom 

 of one of the sides of the basins whose waters require to be renewed, 

 as the Gulf-stream occupies the east side of the North Atlantic. It 

 is plain that the cold and hot waters of two regions can be exchanged 

 by the latter passing underneath the former ; and although the arctic 

 current from the Greenland sea does not contain much ice to the 

 southward of Cape Farewell, it is more than probable its chilly waters 

 pass over a fork of the Gulf-stream, which ultimately sweeps along 

 the shores of "West Greenland. 



May 26. — The following communications were read : — 



1. A letter from Mr. Joule to Colonel Sabine, Treas. V.P.R.S. &c. 



Acton Square, Salford, May 23, 1853. 



My dear Sie, — I notice in the Proceedings of the Royal Society 

 for April 21, a letter from M. Regnault in which some experiments 

 of my own are referred to in a manner which I feel does me injus- 

 tice. M. Regnault says, " Le nombre trouve par M. Joule pour la 

 chaleur specifique de l'air sous pression constante (0*226) est beau- 

 coup trop faible. Celui qui resulte de mes experiences tres nom- 

 breuses, et faites dans des circonstances variees, afin de reconnaitre 

 et d'eliminer les erreurs constantes, est 0*23 7." 



Now, in my paper on the air-engine, Phil. Trans. 1852, part i. 

 p. 74, I have given the results of three series of experiments, viz. 

 0*23008, 0*22674, and 0*2325, and remark, " The mean of the three 



