594 
MR. J. PRESTWIC1I ON SUBMARINE TEMPERATURES. 
two exceeding 4000 feet; and these were made at some distance from those of the 
French observers, who experimented chiefly between Norway and lat. 76° N., whereas 
Scoresby’s observations were mostly north of that latitude, and in the sea west of Spitz- 
bergen as far as 80° north. 
It is easy to determine the depth at which, in inland seas like the Mediterranean, the 
effect of the diurnal variation of temperature ceases, but it is a much more difficult 
problem in Arctic seas. Exposed to the low temperatures of an Arctic climate, the 
surface-waters may continue to sink until their temperature is reduced to 25 0, 4, the 
point at which they attain their maximum of density. This, however, can only happen 
in a state of perfect calm or with waters of unusual saltness, as sea-water of the usual 
specific gravity freezes under ordinary conditions at 27°"4 F., though it has been shown 
that in a state of perfect rest it may be reduced to 20°, or even lower before freezing. 
Under these conditions, and with the complicated action of warm currents from the 
south and of cold currents from the north, we must expect to find considerable variation 
in the temperature of the Arctic Ocean, down, at all events, to the depths hitherto reached 
of 4600 feet. Judging from the conditions prevailing in the Antarctic seas and the sea 
of Baffin Bay, it seems probable that more uniform readings will be obtained at greater 
depths, and that the anomalous readings in the upper strata are caused by the warmer 
waters which flow in from the south tending to take at and near the surface the tempe- 
rature of the air at different seasons, while the deeper part of this mass of warm water 
remains unaffected ; and in the deeper channels there may be, flowing from the north, 
the more permanent body of cold water produced by the winter refrigeration of the 
polar seas of still higher latitudes. 
Subject to the corrections for the causes before named, Scoresby’s experiments 
command confidence. The effect of the corrections will be to reduce his readings where 
Six’s thermometer was used, while where the water-bucket alone was used a small 
addition may be generally needed. 
In the mean time (1815-18) another Russian voyage of circumnavigation*, under the 
command of Otto von Kotzebue, was undertaken for scientific purposes. One hundred 
and sixteen carefully conducted experiments (often taken from day to day) were made 
in both the great oceans and amongst the islands of the Eastern Archipelago. These 
observations, many of them serial, taken at depths of from 24 to 2448 feet, were tabu- 
lated in the order of date. On this voyage Kotzebue used English-made (Jones) Six’s 
thermometers. They were protected by a wooden case closed with “ wire grating,” but 
not in any other way, and they were fastened on the sounding-line about 6 feet above 
the weight. Kotzebue considered that “ seven or eight minutes suffice to give it the 
temperature of the surrounding water, and a quick or a slow pulling up has no effect 
on the observation” (vol. i. p. 89). 
* Entdeckungs-Reise in die Siid-See nnd nach der Berings-Strasse zur Erforschung einer nordostliehen 
Durchfahrt auf dem Schiffe Rurick (Weimar, 1821), dritter Band, von dem Naturforscher der Expedition 
Dr. Chamisso, Tables, p. 230 ; and Dr. Horner’s Report thereon, p. 233. 
