476 
Proceedings of the Royal Society of Edinburgh. [Sess. 
e.g. A change in pressure of 150 gives a change of temperature which 
is of the same order as the possible error in the measurement of the 
temperature. The results given in above Table were obtained with a 
dilatometer of the shape in fig. Sc, and I quote here a few more results 
got by the same method : — 
Pressure = 
300. 
400. 
500. 
600. 
700. 
750. 
800. 
1000. 
f 
Corresponding j 
Temperatures, j 
l 
320 
318 
335 
352 
358 
355 
348 
426 
505 
335 
385 
378 
340 
410 
357 
445 
710 
395 
549 
360 
460 
430 
430 
426 
658 
905 
375 
500 
These corresponding temperatures mean that the water had the same 
volume for the various pressures at the temperatures given in the horizontal 
rows. The values of the volumes for the several rows could not be 
accurately determined ; but by taking the values for pressure 1000 we 
could get the corresponding points on the curves in fig. 4, and might thus 
draw in with fair accuracy the isopiestic for 500 at any rate up to 
about 400°. 
Referring back to the table of critical values on page 475 I should like 
to add the following remark. The method employed by Nadejdine in his 
determinations, using as he did a differential densimeter, was such that 
the critical temperature was observed directly, while Battelli’s values were 
indirectly obtained in the sense that his method was a graphical one. That 
the latter method, however, is a reliable one has been shown by Amagat’s 
results for the isotherms of C0 2 and also by Ramsay and Young’s results 
for other substances. If in fig. 4 we draw in the isotherm for 365° (shown 
by a broken line in the figure), we see that it cuts the 400 isopiestic at a 
point whose abscissa is 320. It must, therefore, cut the isopiestic for 200, 
or the critical isopiestic at a point for which Y is greater than 320. It is 
clear, then, that Nadejdine’s value for the crit. vol. — viz. 2*3, marked ^ in 
figure — is inconsistent with our results ; while the value 4*8, given by 
Battelli, marked x , seems quite reasonable. The two graphical methods, 
therefore, agree in placing the critical volume nearer 4*8, than 2*3. 
The results obtained in this paper, then, refer to water under pressures 
above the critical pressure. The range of temperature for all pressures 
investigated includes the critical temperature. It is clear from the figure 
