HYDROGRAPHY. 



83 



find S by multiplication with the so-called coefficients of specific-gravity and chlorine. Many naturalists 

 have tried to determine these coefficients, and found values deviating so much from one another, that 

 the use of any of them, in so far as ocean-water is concerned, would easily give values for S which 

 differ 0-06 %o from each other, and thus have the effect of making it impossible to make a direct 

 comparison between the results of the different expeditions. Nearly all the observers have as a basis 

 for the calculation used the determination of the amount of chlorine executed by volumetric titration 

 with a solution of silver-nitrate and chromate of potassium as index, as this way of determination 

 gives the greatest exactness in proportion to the work required. 



The question is now: What is obtained by going from the amount of chlorine to the total amount 

 of salt, and what is upon the whole the intention with the determination of chlorine and salinity? 



The latter question may be answered this way that it is desired to find a number, by which 

 each water-sample is characterised. The amount of chlorine must, however, be able to do this as well 

 as the total quantity of salt computed by the amount of chlorine. If we set down the quantity of 

 salt as the principal feature of a water-sample, we shall only attain to bring about an incongruity 

 between different series of observations, as it may be disputed which coefficient is to be used, and in 

 case of a direct determination of the total quantity of salt, we shall run a still greater risk, as the 

 probable error arising from this, will be at least as great as the one resulting from a determination of 

 chlorine by volumetric titration, and far greater than the error appearing in a determination of chlorine 

 by weight- titration. It seems as if the experimentalists (Ekman, Forsberg) have come to the con- 

 clusion that the coefficient of chlorine is increasing with a decreasing quantity of salt But this will 

 be entirely trifling in proportion to the variation of the salinity, from which follows that the coeffici- 

 ents will not, when used in this manner, contribute to characterise a given sample of water in pro- 

 portion to others. 



It will be quite different, if for a single sample of water we consider the amount of chlorine 

 as the principal feature, in which case the determination of the coefficient of chlorine will be able to 

 give a new characterising cipher for the water-sample. It is evident that the coefficient of chlorine 

 of a water-sample does not alter, even if the sample be mixed with pure water, and when it is found 

 that water of little saltness has a great coefficient of chlorine, this is very likely not to be understood 

 otherwise than that the generation of such water is due to the mixture of salt water with a little 

 coefficient of chlorine, with fresh water with a great coefficient, such as it is the case on the coasts 

 and in the seas where the influence of water from the rivers can be traced. Whether anywhere in 

 nature there is a mixture between sea-water and pure water, must be considered as doubtful; still this 

 must to a certain degree be the case under rain-fall and melting of icebergs, and at such places we 

 may then expect a normal coefficient (the one which applies to ocean-water). 



As, however, the uncertain determination of the total quantity of salt, is inseparable from the 



determination of a coefficient of chlorine, it may be considered as doubtful how far this determination 



is worth the trouble attending it. One would think that by other proceedings it might be possible 



to have stated for certain, and with greater exactness, whether a trifling salinity of the sea-water is 



due to the mixture of sea-water with fresh water from the rivers, or with water produced by the 



melting of the glacier-ice. 



11* 



