162 



Mr. J. Y. Buchanan on the Use of [June 15, 



tigated for the soundings in question, the weight of a cokimn of sea- 

 water in any of these localities can be calculated with great accuracy. 



The observations which have been taken as a basis for determinations 

 of depth were made in the latter part of the year 1875 in the South- 

 Pacific Ocean. They were twenty in number, and were made at depths 

 varying from 500 to 2300 fathoms, and at temperatures varying from l°-4 

 to 4°*03 C. The mean compressibility of water determined from these 

 observations was 0*0008986 per 100 fathoms of sea-water, the extreme 

 values being 0*000915 and 0"000882. Observations made at greater 

 depths in the North Pacific gave as a mean of six observations at depths 

 varying from 2740 to 3125 fathoms the value 0-000878, indicating a 

 slight diminution in the coefficient of compression at very high pres- 

 sures. 



The effect of pressure being thus known, we are in a position, by 

 comparing the indications of the instrument with those of a trustworthy 

 deep-sea thermometer, to determine the absolute depth to which it has 

 been sunk beneath the surface ; and assuming the depth as indicated by 

 the sounding-line to be correct, we should be able to determine the tem- 

 perature at the depth in question from the indications of our instrument, 

 and without the use of a thermometer. Por the latter purpose, however, 

 the instrument, as above described, is useless, because the dilatability of 

 water at the low temperatures obtaining in deep water is so small as to 

 be negligible compared with its elasticity. 



The application, however, of the principle above indicated would mani- 

 festly present some very great advantages in the determination of deep- 

 sea temperatures. 



In the open ocean, where, as a rule, the temperature diminishes con- 

 stantly as the depth increases, the Millar- Gasella thermometer gives 

 sufficiently accurate results. In the case of enclosed seas, or in the 

 neighbourhood of ice, however, this is not always the case. In the Medi- 

 terranean, the Eed Sea, and many of the seas of the Eastern archipelago, 

 besides, possibly, large tracts both of the Atlantic and Pacific Oceans, the 

 temperature decreases regularly down to a certain depth, which is different 

 for different seas ; and at all greater depths the Millar- Casella thermo- 

 meter gives identical readings, indicating that the water is either at the 

 same temperature or some higher one. In the neighbourhood of ice, layers 

 of water are frequently met with at various depths whose temperature, 

 being higher than that of the surface, is indicated by the maximum index of 

 the Millar-Casella thermometer. Besides these layers there may be, and 

 there probably are, others whose temperature is higher than that of the 

 water immediately above them without reaching that of the surface, and 

 their temperature would remain unrecorded. It would therefore be of 

 great advantage if the piezometer could be adapted for the determination 

 of temperatures at known depths. An efficient instrument for this pur- 

 pose has been obtained by filling the bulb of the piezometer with mercury 



