PHYSICAL PROPERTIES OF WATER, ETC. 43 



IX. Compressibility, Expansibility, etc., of Solutions of 



Common Salt. 



This part of the inquiry was a natural extension of the observations on sea-water, 

 but it was also in part suggested by the fact that an admixture of salt with water 

 produces effects very similar to those of pressure. Thus it appeared to me that an 

 investigation of the compressibility of brines of various strengths might throw some 

 light on the nature of solution ; and also on the question of the internal pressure 

 of liquids, which (in some theories of capillary forces) is regarded as a very large 

 quantity. 



The solutions experimented on contained, roughly, 4, 9, 13"4, and 17'6 per cent, 

 of common salt. The piezometers used for the experiments already described were 

 filled with these solutions in July 1887; one, for comparison, being left full of 

 fresh water. I obtained a large number of results at temperatures about 1°, 9°, and 

 19° C, and at 1, 2, and 3 tons weight per square inch. Unfortunately these were still 

 more discordant than those made with sea- water ; so much so, in fact, that an error of 

 1 or occasionally even 2 per cent, was not by any means uncommon. However, by 

 plotting all the observations exactly as described in the two last sections, I found that 

 they could be fairly represented by the curves shown in Plate I. In most cases two 

 at least of the three points for each curve were fairly determinate ; one of these being, 

 in all cases, within a degree or so of 10° C. For this was obtained by experiments in 

 the large gun, where the difficulty of relieving the pressure without jerks is much less 

 than in the smaller apparatus. Of the general accuracy of these curves I have no 

 doubt. Thus, for instance, it is certain that the compressibility at any one temperature 

 and pressure diminishes rapidly as the percentage of salt increases. And the rate at 

 which the compressibility (for any one range of pressure) diminishes as temperature 

 increases, becomes rapidly less as the solution is stronger. My observations do not 

 enable me to settle the more delicate question of the variation of the rate at which the 

 compressibility (at any one temperature) falls off with increase of pressure in the 

 various solutions. For the limits of error in the various determinations, especially 

 with the more nearly saturated solutions, are quite sufficient to mask an effect of this 

 kind unless it were considerable. An attempt, however, will be made in next Section. 



There is little to be gained by putting the results of the inquiry in a tabular 

 form ; for they can be obtained from the plate quite as accurately as is warranted by 

 the limits of uncertainty of the experiments. See p. 47. 



I am indebted to Dr. Gibson for the following determinations, which have a high 



