A Retrospect of Oceanography 



tions are likely to be made with equal frequency above that 

 temperature and below it, 19-6 C. would appear to be a suitable 

 temperature for general reduction. At the same time, the 

 area of sea surface having a higher temperature than this 

 is much greater than that having a lower temperature, and 

 as the surface temperature of the sea is the principal factor 

 determining the temperature of the air in a ship, it would seem 

 reasonable to put the common temperature rather higher 

 than lower. I have gone over the list of density determina- 

 tions of surface water in the "Challenger," and only eighteen 

 were made at temperatures above 29-0 C. in the seas round 

 New Guinea. Subsequent experience in the "Buccaneer" in 

 the Gulf of Guinea and on a passage from Panama, across 

 the corresponding homologous feature of the Pacific, the great 

 Central American bight, to San Francisco, showed that in 

 these parts of the ocean during one-half of the year the 

 temperature of the surface water and of the air is generally 

 above 29 C. 



In the following table are given the mean temperatures 

 at which the density of surface water was determined in the 

 different oceans: 



