SEA-SALT AND GEOLOGIC TIME 57 



York State, where brine-springs are plentiful, there is no 

 appreciable effect on the salt content of the rivers. 1 The only 

 known means by which fresh chlorine reaches the sea is volcanic 

 action, and it is a point open to dispute how much of the volcanic 

 chlorine is not ultimately derived from the sea. It follows, 

 therefore, that, of the sodium which actually reaches the sea, 

 only that not associated with the chlorine can be counted. 



This much Prof. Joly and those who agree with his earlier 



estimate have been willing to admit. But Prof. Joly maintains 



that, if the chlorine equivalent of the sodium be subtracted, 



there is still sufficient sodium to necessitate an estimate of 



geologic time less than 150 millions of years. His reason is 



merely Sir John Murray's rough tabulation of then current 



analyses and some more recent results. It does seem strange, 



however, that Prof. Joly never troubled to inquire whether there 



were any water analyses sufficiently accurate for his purpose. 



It is highly probable that Prof. Joly's original paper would 



never have been written if he had understood why water 



analyses are undertaken, and the manner in which they are 



actually performed. Had he been a water analyst, or even a 



chemist, the first thing that would have occurred to him would 



have been that these sodium determinations were decidedly 



hypothetical. Several chemists have expressed doubts as to the 



validity, but such discussions Prof. Joly has either ignored or 



failed to understand. 2 It may, therefore, surprise Prof. Joly 



to be informed that it is doubtful whether the sodium content of 



any single river water has ever been accurately determined. If 



any such cases have occurred, they are very few. Let us 



imagine that there is, in a given sample of river water, two 



parts of sodium per million. Such a proportion would be quite 



ordinary according to the usual tables. It would be a very 



interesting problem to try to separate this out and weigh it. To 



obtain a good weighable quantity (say '05 gram of sodium giving 



about *i 5 gram of sodium sulphate 3 ) would require 25 litres of 



1 For further information on these points see Jackson's " Normal Distribution 

 of Chlorine," U.S.A. Geological Survey, Water Supply Paper No. 144. 



2 See particularly discussion with Mr. Acroyd, Chemical News, 1901, and F. W. 

 Clarke, Data of Geochemistry, p. no. 



3 In ordinary accurate analysis sodium is usually weighed as sulphate. In 

 water analysis, however, the quantity is so small that conversion to sulphate is not 

 worth while. The residue is reckoned as chloride, though it need not necessarily 

 be so. 



