SUNIER : Marine fish-ponds of Batavia. 169 
These salts coming from the fresh water will, however, be included 
to the extent of only less than 6 °/) of their real amount in the salinity 
stated to correspond with the observed amount of chlorine in KNUDSEN’s 
tables ('°). For, whereas the salts dissolved in the sea-water belong for 88'/, 0/, 
to the chlorides, 11 °/, to the sulphates and only 0.3 °/, to the carbonates, 
the salts contained in fresh water on the contrary consist for 60 °/, 
of carbonates, 10 °/, of sulphates and only just over 5 °/, of chlorides. 
The proportion of chlorides expressed in the total amount of the salts 
in fresh water is therefore less than 6 °/, of the corresponding proportion 
in sea-water. 
However the influence of the salts derived from the fresh water on the 
salinity of the pond-water cannot be great, as will appear from the following 
calculation. 
A salinity of 7!/,®/, would result from the mixture of 1 K. G. of 
sea-water of a salinity of 30°/,, with 3 K. G. of distilled water. On mixing 
one K. G. of sea-water of a salinity of 30°/)) with three K. G. of 
river-water, the salinity of which we may put at0.2°/,, at the outside, the 
salinity of the mixture would then work out ype Olio = 108 ews 
assuming that in the admixture no salts disappear from the solution by 
precipitation. 
This difference of 0.15 °/o) in the real salinity of the two mixtures 
would therefore in this case — apart from possible errors in determining 
the amount of chlorine — correspond to a difference of not quite 6°/, of 
0.15 °/o,, that is not even 0.009 °/,, of the salinity to be derived from the 
amount of chlorine observed. On the other hand — again apart from in- 
accuracies in the areometer and thermometer readings — the difference of 
0.15 ©}, would be expressed to practically, though not quite the correct 
amount in the salinity calculated by the KNUDSEN-tables (!°) from areometrical 
and thermometrical observations. 
In both cases however the respective differences of 0.009 °/,, and 
0.15 Jo), are smaller or at least not greater than the average error that 
may be expected in applying either method. 
In connection with the combined influence of the organic matter 
dissolved in the pond-water, the oozy and detritus-particles in suspension, 
and any dissolved salts carried by the fresh water, the figures for the salinity 
of fish-pond water, determined by areometer and thermometer readings, may 
however be expected to be generally higher than those resulting from the 
determination of the amount of chlorine. 
For the purpose of checking the correctness of this surmise I requested 
our hydrographic assistant, Mr. K. M. VAN WEEL, to determine the amount 
of chlorine in ten samples of fish-pond water, of which I had myself deter- 
mined the specific gravities by means of areometer and thermometer 
readings. The result may be tabulated thus: 
