THE PROPERTIES ÔF SEA-WATER 7i 



The quantity of dissolved salts in sea-water can be given 

 approximately. Assuming the average density of sea-water, 

 with an allowance for increase in the lower layers due to 

 pressure, to be i"04, the total weight of the seas would be 

 138 X 10^^ metric tons. Allowing the weight of sait to be on 

 the average 3J per cent, of this, the total weight of dissolved 

 salts is 4-84x10^^ metric tons. The volume of sait which 

 would remain if the waters of ail the océans and seas of the 

 world were evaporated dépends on the spécifie gravity of the 

 salts, which may be taken, on an average, to be 2-22. The 

 volume of the sait would consequently be 2'i8xio^^ cubic 

 mètres, which if spread out on the floor of the océan would 

 give a depth for the sait layer of igôf feet, nearly 33 fathoms. 

 This mass of sait is considerably more than the whole land 

 volume of Africa (including Madagascar) above sea-level, or 

 three times as great as the volume of Europe, or nearly half 

 that of Asia. 



The density of sea-water is of considérable importance 

 from an oceanographical standpoint, since a great many océan 

 currents are clearly due to différences in the density of the 

 layers of surface water. 



The détermination of the density of sea-water in a physical 

 laboratory is not a particularly difficult opération, and full 

 instructions for determining density are given in the varions 

 textbooks on practical physics. At sea, however, an instru- 

 ment known as an aerometer is used. This is a hydrometer 

 of glass, which, except for the long graduated stem, is com- 

 pletely immersed in the sea-water the density of which is 

 desired. Glass, although more brittle than métal, has two 

 advantages : it is less susceptible to increase in volume owing 

 to changes of température, and it cannot be indented by rough 

 handling. Since it is only desired to read densities from 1,000 

 to 1,031, it is possible to construct a scale for an aerometer to 

 function between thèse densities. In order to allow of minute 

 readings the scale would hâve to be a very long one, so it is 

 found advisable to hâve a séries of aerometers, usually six. 



