236 DEPTHS OF THE OCEAN 



The numerous other substances in solution are present in 

 such extremely small quantities that they may be disregarded. 

 Although the total salinity may vary widely, the composition of 

 the dissolved solids proves to be practically the same every- 

 where. Hence if in a sea-water the percentage of any one 

 component, say chlorine, be known, the total salinity can be 

 ascertained by calculation. 



The direct determination of salinity by evaporating a known 

 volume of water to dryness does not give accurate results, unless 

 the amount of chlorine is carefully determined before and after 

 the evaporation, because in the last stages of evaporation and 

 in drying the residual salt uncertain amounts of chlorine are dis- 

 engaged in the form of hydrochloric acid. Such a determination 

 is very circumstantial, and it is therefore necessary to resort to 

 indirect methods, which may be physical or chemical. 



An old-established physical method consists in determining 

 the density by means of the hydrometer. This is a glass cylinder 

 which floats in the water and has a graduated stem, on the scale 

 of which densities are read off The temperature of the water 

 must be determined at the same time. Densities so found are 

 recalculated by means of tables to a standard temperature, 

 generally 17.5° C. Now, owing to the uniform composition 

 of sea -salts, a definite density at 17.5° corresponds rigidly 

 to a definite salinity. Hence by referring to tables the 

 salinity of a sea-water can be found from its density at standard 

 temperature. 



The hydrometric method is easily applied on board ship, 

 and may be made to give densities correct to four places of 

 decimals. Densities can be determined to a yet higher degree 

 of accuracy by means of the pycnometer, but this method is 

 practicable only in a laboratory on land, and is not often 

 employed. 



Two other physical methods have been tried by way of 



errors have been detected in earlier determinations referable to the leaky condition of the 

 water-bottles. 



When the forms of apparatus described above are to be used, the vessel must be stopped and 

 hove to as long as the work goes on. Recently several investigators have studied the problem 

 of constructing a i apparatus to be used while the ship is under way. Water-bottles have been 

 made which can be let out when the ship is going at full speed, with the line running'freely so 

 as to allow them to sink. On checking the line the apparatus is closed by a mechanism like 

 that used by Buchanan in his water-bottle. The water-bottle being insulating, a temperature- 

 reading is secured together with the water-sample. In such an experiment a metre-wheel 

 showing how much line has run out is no use ; one must have a special depth-gauge, usually 

 one to measure the compression suffered by a certain volume of air from the weight of the water. 

 These new instruments are not in common use as yet, being still in the experimental stage, 

 but the time is not far off when we shall have automatic water-bottles working with absolute 

 precision. That will mark an important step forward, as much time will then be saved in 

 an expedition. 



