374 FETTER8S0N, SIBERIAN SEA. 



water beneath the ice was for tlie most part cooled below its 

 normal freezing point [ — 1°.5 to — 1°.? C]. From the increasing 

 thickness of the ice-floes we must infer, that the sea-water 

 below was constantly giving up ice. Consequently we ought 

 to expect, that its temperature would remain precisely at the 

 normal freezing-point during the whole time. Althongh I am 

 acquainted with many similar observations from previous 

 deep-soundings in arctic seas, I can find no plausible explana- 

 tion for the stränge fact, that water, which is in contact with 

 ice, can be over-cooled at the same time. From a physical 

 point of view such temperatures as f. ex. — 2°.8 C or — 3°.o C 

 in water, which does not contain more than 2.72 p. c. of salt 

 and has a spec. weight of not more than I.0217 — 1.0227 or 

 thereabouts must be considered as abnorm al. 



The lowest temperatures hitherto observed in the ocean 

 are those found in Baffin Bay [t = — 2°.8 C; depth = 1478 

 metres (Parry), and t = — 3°.7 C; depth = 1243 metres (Ros s 

 & Sabine) and west of Spitzbergen, 1873, b}" the Swedish 

 expedition [t = — 3°.2 C; depth = 142 metres] and (in the same 

 year) by Mr. Leigh Smith [t ^ — 5°.i C; depth = 1005 metres]. 

 The numbers in the above tables, taken from the observations» 

 in the winter 1878 — 79 at Pitlekaj, do not range among the 

 lowest submarine temperatures ever observed, hut are, however, 

 still more remarkable by the circumstance, that the cold layer 

 was found only a few metres below the surface. The low 

 temperatures observed in the sea east and west of Greenland 

 belonged to a stratum of water 140 to 1000 metres beneath 

 the surface, which was exposed to an enormous pressure. 

 The råte, at which the freezing point of salt water decreases 

 under the influence of pressure, is at present totally unknown 

 to US. It might therefore be possible that the water of — 5° C 



^ We ought by no means a priori to apply the formula found by J. 

 Thomson (Transact. R. S. E. T. XVI) for the depression of the freezing- 

 point of pure water also to sea-water. Tlie mistakes, which have arisen 

 from confounding the properties of sea-water and fresh water, or sea-ice with 

 fresh-water-ice, are too numerous and flagrant not to impress the necessity 

 of experimental investigation. In the same way, as the majority of hydro- 

 graphers 30 or 40 years ago took for granted, that the maximum density of 

 sea-water was at + 4° O, like that of fresh water, and that the sea from a 

 thousand metres from the surface to the bottom had a constant tempera- 

 ture etc, Science has hitherto tacitly supposed, that the dilatation, the chem- 

 ical composition and the latent heat of sea-ice was equal or almost equal 

 to that of fresh-water-ice. In the preceding paper I have tried to show up 

 the fallacy of these assumptions. 



