10 FOFONOFF [chap. 1 



collected from a depth of 3000 m off the coast of Portugal. Part of the sample 

 was diluted with distilled water to a salinity of 3 1.13 %o, and the rest evaporated 

 to a salinity of 38.83 %o. Although he carried out some measurements of com- 

 pression by lowering sea-water samples into the ocean to subject them to 

 pressure, his final results were based entirely on subsequent determinations in 

 the laboratory. Pressures in his laboratory apparatus were determined by 

 measuring the compression of distilled water at 0°C, and then calculating 

 pressure using Agamat's (1893) determinations of the compression of pure 

 water. 



Ekman measured sea-water compression at several temperatures from 

 0-20°C for pressures of 200, 400 and 600 bars. He used his results to develop an 

 empirical formula for the mean compression of sea-water, fx, in the form 



ju, = {ao-a)lpao, 



where a is the specific volume ^ at pressure p and ao the specific volume at 

 atmospheric pressure. Specific volume at pressures above atmospheric pressure 

 is computed from 



a = ao{\-fxp), (25) 



where 



lOV = [4886/(1 + 1.83 X 10-5j9)]- (227 + 28.33r?-0.551j^2 + o.004j^3) 



+ 10-4p(105.5-f 9.50?^- 0.158t?2)_ 1,5 x IQ-^^p^- 



- 10-i(cto - 28)[(147.3 - 2.T2& + 0M&^) 



- 10-4p(32.4 - 0.87i^ + 0.02j^2)] 



+ 10-2(cto - 28)2[4.5 - O.lj? - 10-4^9(1.8 - 0.0Q&)] (26) 



and^ is pressure 2 in decibars (10^ dynes/cm2).3 



Eckart (1958), in a careful study of the density and compression of pure- and 

 sea-water, has concluded that the equation of state for both can be represented 

 with sufficient accuracy by the Tumlirz equation of the form 



{p+po){a-ao) = A, (27) 



where A= 1779.5+ 11. 25z^-0.0745i^2_(3,804-0.01i^)*S, ao = 0.6980, ;po = 5890 + 

 SSd' — O.SlB&^ + SS, p is total pressure in atmospheres (1 atm= 10.1325 db), and 

 a is specific volume in millilitres per gram. 



1 The specific volume, a, is given in millilitres per gram and is equal numerically to the 

 reciprocal of specific gravity. To obtain specific volume v in cubic centimetres per gram, 

 a is multiplied by 1.000027 cm^/ml. 



2 The pressure, p, is total pressure less one standard atmosphere. This convention is 

 widely used by oceanographers. 



3 Bjerknes and Sandstrom (1910) gave the coefficient 0.002 instead of 0.02 for the 

 (ao — 28)p&^ term in (26). The discrepancy is probably due to a misprint. However, it is 

 not clear in which paper the misprint occurred. The difference between the two formulae 

 appears to be completely negligible for ranges of variables encountered in the ocean. 



