1912-13.] Compressibilities of Solutions of Inorganic Salts. 283 
to take a few salts and examine a number of dilute solutions of these at 
pressures at least up to 1000 atmospheres. These would allow me, in general, 
to obtain an idea of the variation of internal pressure with concentration and 
pressure, and in particular to test whether the formula from which Tammann 
calculates the internal pressures is satisfied in these particular cases. 
The substances chosen were Na 2 C0 3 , NaOH, KOH, MgS0 4 , and ZnS0 4 . 
I measured the densities and compressibilities at 15° C. of five solutions of 
NaOH, and four solutions each of Na 2 C0 3 , KOH, ZnS0 4 , and MgS0 4 , making 
twenty-one solutions in all, all of normal concentration or under. A series 
for water was also taken. 
The substances were the purest obtainable from Kahlbaum. They were 
in each case assumed to be chemically pure without further chemical 
examination. The sodium carbonate was prepared in the usual way from 
sodium bicarbonate. The water was commercial distilled water, redistilled 
in Jena glass vessels, the middle portion being chosen. The flasks used 
for holding and preparing the solutions were also of Jena glass, and had 
been previously calibrated. The general procedure was to make in each 
case an approximately normal solution and to dilute this as required. The 
whole process of solution and dilution was carried out in a constant- 
temperature bath. The only solutions which stood for any length of time 
(of course in well--stoppered flasks) were the normal solutions. The other 
solutions were prepared just before use. 
The density was measured by means of an ordinary pyknometer kept 
for hours in a constant-temperature bath at 15° C. This method, though 
perhaps not so reliable for very dilute solutions as the upthrust method 
of Kohlrausch, gave me sufficiently accurate results for my purpose. A 
mean of three observations was in general taken. The weights used had 
been previously calibrated, and care was taken that the balance temperature 
did not vary much from 15° C., and that otherwise evaporation did not 
affect the results. 
The compression cylinder used was of the same pattern as that described 
by me in a previous communication,* the difference being that I did not 
require the arrangement for high-temperature work. The cylinder itself 
was made under my directions in Leipzig, and the plug arrangement for 
introducing the wires was completed by Herr Dornfeld, Mechaniker in 
Professor Des Coudres’ laboratory. The pump, compression screw, and 
gauges were obtained from Schaeffer & Budenberg. The gauges were 
two in number, one reading by kgs. up to 400 kgs./cm. 2 , the other reading by 
10 kgs. up to 3000 kgs./cm. 2 In general one could interpolate to r V of a 
* Proc. Roy. Soc. Edin 1911 . 
