758 
Proceedings of the Royal Soeiety 
where t is the temperature centigrade, and p the pressure in tons 
weight per square inch. This, of course, cannot he the true 
formula, hut it is sufficient for ordinary purposes within the limits 
of temperature and pressure above stated. It represents the value of 
With a new set of compression apparatus, very much larger and 
more sensitive than those employed in the above research, I have 
just obtained the following mean values for the single temperature 
15° -5 C. 
Pressure in Tons. 
1. 
1|-. 
2. 
3. 
Fresh water, . 
0-00678 
663 
657 
638 
Sea water, 
0-00627 
618 
609 
593 
These are the values of ^ ; and they give, for the true com- 
pv^ 
pressibility at any pressure, and temperature 15° *5 C., 
the formulae. 
Freshwater, . . 0*00698(1 - 0*05 p) 
Seawater, . . . 0*00645(1 -0*05^) 
The ratio is 0*925, i.e., the compressibility of sea water at the 
above temperature is only 92*5 per cent, of that of fresh water. 
\_Added 28th October 1884.] 
With the notation employed {ante,^. 229), ^^^ = 0*000038/152 
p being in atmospheres ; while we have e = {t - 4)/72,000. Hence 
U 0*000038. - 
Thus, for hp = \52 atmospheres, i.e., 1 ton weight per square inch, 
we have 
U= -2°*74 C. 
This agrees, in a remarkable and altogether unexpected manner, with 
the result 2° *7 C. obtained by direct measurement {ante, p. 228). 
