ATLAN1. DEEP-SEA EXPED. 1910. VOL. i] PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



65 



SB 27 B 

 350 2 



1400 



2 260 258 



i3° i4' 15' 



Fig. 23. Comparison between tlie "Micliael Sars" stat 

 "Tlior" station 180 (in 1906) 



ion 92 (in 1920) at 18° 29' N., 13° 55' W. and the 

 at 48° 19' N., 13° 53' W. 



the temperature of the water in the absence of any foreign 

 bodies. 



Let us assume that the water from the surface to a 

 great depth is so perfectly inixed that a neutral equili- 

 brium is established. The salinity must then be the same 

 at all levels, while the temperature will show an increase 

 downwards corresponding to adiabatic changes. The 

 "bottom-water" in the great depths of the oceans is homo- 

 haline, and is relatively cold. It must have been at or 

 near the surface in some region where it obtained a low 

 temperature before sinking to the abysses of the sea. 

 While sinking it is heated adiabatically. If the deep 

 water consists of water-masses which originally — when 

 leaving the upper levels — were perfectly uniform with 



regard both to salinity and temperature, and the uni- 

 formity has not been disturbed by other processes than 

 variations in pressure, or if the descending water-masses 

 are thoroughly mixed at greater depths, we may expect 

 to find an adiabatic distribution of temperature. A slow 

 heating takes place from the bottoin of the sea and con- 

 tributes to create such a distribution; in certain circum- 

 stances we may even expect an "over-adiabatic" fall of 

 temperature from the bottom upwards. 



If the salinity is uniform and the vertical distribution 

 of temperature is isothermal, there will be a state of 

 positive stability. 



With uniform salinity and an adiabatic distribution 

 of temperature the value of </, decreases slightly down- 



