572 
Proceedings of the Royal Society 
3. On the Accurate Measurement of High Pressures. 
By Professor Tait. 
In the course of an examination of some of the “ Challenger ” 
deep-sea thermometers, I have recently had occasion for measure- 
ments, accurate to one or two per cent., of pressures such as five or 
six tons weight per square inch. The ordinary gauges showed 
themselves to he quite untrustworthy, and it was necessary to devise 
some plan of whose accuracy the experimenter can feel assured. 
The following process has proved completely successful, and is 
capable of any desired degree of accuracy. 
Simple methods based on the compression of gases, such as air or 
nitrogen, are of the highest value wherever they can he adopted ; 
for the law of compression of these bodies is known with great 
accuracy (at least for one definite temperature) from the measure- 
ments recently made by Amagat, in which the pressures were 
directly reckoned in terms of a column of mercury. A simple form 
of gauge, in which the column of mercury compressing the gas into 
a small bulb at the extremity is made to break off at a constriction 
in the connecting tube, enabling us (by weighing the mercury forced 
over into the bulb) to measure the compression very accurately, 
suffices amply for all pressures up to a ton weight per square inch, 
or even farther. 
But this instrument becomes rapidly less and less sensitive at 
higher pressures ; so that though the law of compression for a con- 
siderably extended range is now known, for pressures above a ton 
something else is required. 
Hooke’s Law now comes to our assistance. An instrument re- 
sembling a thermometer in form (but with a tube of much larger 
section compared with the capacity of the bulb than is usual in 
mercury thermometers) supplies the next step. It is filled with 
mercury (because of the small expansibility of that liquid), and is 
thus practically unaffected by small changes of temperature. Over 
the mercury in the stem is a long column of alcohol in which the 
index moves, and the rest of the tube contains alcohol vapour only. 
The bulb is made cylindrical for several reasons ; the chief being 
to secure uniformity of thickness, which is practically unattainable 
