64 Proceedings of Royal Society of Edinburgh. [sess. 
These considerations seemed to point to the quantity B as being 
at least closely’’ connected with the internal molecular pressure 
(usually named after Laplace) ; and, speculative as the idea con- 
fessedly is, it seemed worthy of further development. Another 
argument in its favour is furnished by a consequence of the hypo- 
thesis. For it is easy to see that when the average compressibility 
of a substance can be represented by the expression above, the 
equation of its isothermals must have the form 
(B + p)(*,-a) = C; 
approximately that given by the kinetic theory of a gas, when it is 
regarded as an assemblage of hard spherical particles. 
iN’early three years ago, while I was preparing for press the second 
edition of my text-book “ Proj)erties of Matter f M. Amagat kindly 
gave me several unpublished numerical details of his magnificent 
experiments on the compressibility of water and ether. The follow- 
ing short table gives in its second column some of these results for 
water at 0° C. : — 
Pressure. 
Yoliime. 
a 
b 
c 
1 
1-00000 
1-00000 0 
1-00000 0 
1-00000 0 
501 
•97668 
•97664+ 4 
•97652 + 16 
•97657 + 11 
1001 
•95645 
•95662 - 17 
•95644+ 1 
•95652- 7 
1501 
•93924 
•93925 - 1 
•93909 + 15 
•93916+ 8 
2001 
•92393 
•92405 - 12 
•92393 0 
•92399 - 6 
2501 
•91065 
•91064+ 1 
•91058+ 7 
•91062+ 3 
3001 
•89869 
•89870- 1 
•89873- 4 
•89875- 6 
The numbers in the columns a, b, c are volumes calculated respec- 
tively from the following formulae for the average compressibility 
for p atmospheres : — 
•30454 -30 -3015 
6019-}-^’ 5887 -i-i?’ 5933 -f-j/ 
The first was calculated from the data for 1, 1501, and 3001 atm. ; 
the second from those for 1, 1001, and 2001 atm . ; the third was 
obtained from them by interpolation. After the numbers in each 
column the difference “ observed - calculated ” is given. These are 
all small ; and, especially in the case of formula c, the coincidence 
