ACKROYD : ON THE CIRCULATION OF SALT. 
411 
or of the load of salt in this special case carried to the sea fully 09-8 
per cent is sea-salt, and only less than 0"2 per cent earth-salt, or that 
part which is to be taken into account in ascertaining the rate of 
solvent denudation now in progress, and even this is cyclic in a sense, 
as it comes from a Carboniferous and goes back to a present-day sea. 
I draw practically the same conclusion from other and more 
general considerations. Thus, the solids dissolved- in average river 
water yield the atomic ratio of sodium to chlorine of 1 : 0-345, as 
calculated from Sir John Murray's data,* while in the Earth's crust, 
using Professor F. W. Clarke's data, we have a ratio of 1 : 0-0024 ;t 
in other words, the chlorine in river water, which we take as a measure 
of the salt going to the sea, exists in 143 times greater proportion 
than we find it in the crust of the Earth, or to put it in yet another 
form probably only 0'7 per cent, of earth-siilt, and 99'3 per cent, 
of salt from some other source, say cyclic salt, is being carried by 
rivers to the sea. 
How do these results square with other facts ] Lawes. Gilbert, 
and Warrington; have shown that the amount of chlorides in drain- 
age from drain-gauges in unmanured land and uncropped soil is 
almost exactly equal to that contained in the rain. In other words 
they were unable to estimate the amount of earth-salt, it was so small. 
Again in chalk, granite, sandstone, and other rocks, the sodium 
chloride is so inconsiderable a trifle as usually to find no place in 
recorded analyses. It is improbable then, that after extended in- 
vestigations, an average of 1 per cent, of the sodium carried to the 
sea as chlorides. tKrc, will ever be found as the result of solvent de- 
nudation, and consequently that Prof. Joly's 90 millions of years 
estimate of the age of the earth based on uniformity of action, would 
become 8,000 millions of years, a quantity of time which I imagine 
will be too excessive even for the exorbitant demands of the biologists. 
Seasonal Variation of Salt Cii^culation. 
In all such problems again seasonal variation of fluviatile salt- 
load must also be taken into account. It has a tendency to average 
*The Scottish Geographical Magazine, Feb., 1887, p. 12. 
tBull, U.S. Geol. Survey, No. 148, p. 13. 
^ Trans. Chem. Soc, Vol. LI., p. 94. 
