220 DEPTHS OF THE OCEAN 



increases, and by this process some heat is taken up which is 

 drawn from the Hquid, lowering its temperature. When, there- 

 fore, a water-sample is drawn up in an insulating water-bottle 

 from a depth of looo metres, the temperature of the water- 

 sample sinks a little. Nansen first called attention to this fact, 

 and has drawn up tables for the corrections according to Lord 

 Kelvin's formula. The corrections prove to be quite consider- 

 able. When employing an insulating water-bottle, account must 

 be taken, not only of the alteration of volume in the water- 

 sample, but also of that taking place in the solid parts of 

 the water-bottle. A water-sample, for instance, brought up 

 in an ordinary-sized Pettersson-Nansen water-bottle from a 

 depth of looo metres in the Norwegian Sea, is cooled 0.06" C. 

 while being hauled up ; a sample from the same depth in the 

 Mediterranean is cooled 0.17' C. This difference is due to the 

 fact that the amount of cooling depends on the temperature of 

 the water, which at 1000 metres in the Norwegian Sea is about 

 — 1° C. and in the Mediterranean +13° C. 



We are here confronted with a problem of considerable 

 interest. When a body of water sinks from the surface down 

 to great depths, its temperature rises a little because of the 

 compression. The " bottom -water " of the Atlantic Ocean 

 averages nearly 2^° C. ; supposing that it has sunk from the 

 surface to a depth of 3000 metres, it has been heated about 

 0.27° C. in the course of its descent, by reason of the increasing 

 pressure. If it should appear at the surface again, the reduc- 

 tion of pressure will have lowered the temperature by the same 

 amount, — 0.27 C. There are various other conditions which 

 produce changes in the temperature, as, for instance, mixing 

 with other bodies of water, in the upper layers absorption of 

 solar heat, near the bottom possibly a very slight influence 

 from the internal heat of the earth. It is, of course, difficult 

 in such a combination of factors to single out the effects of one 

 of them individually. 



During the "Michael Sars " Expedition in the North 

 Atlantic we made a certain number of observations in the 

 deeper layers with a Richter reversing thermometer, which 

 seemed to prove in several cases that the temperature increased 

 slightly towards the bottom. The following extract from the 

 " Michael Sars " tables shows the number of the station, the 

 depth, the temperature (measured in szhi), and the temperature 

 that the water would acquire — on account of the reduction 

 of pressure — if it were raised to the surface. The latter 



