( 948 ) 
upper portion of the manometer tube and the tube of the volumeno- 
meter. by this means control measurements are possible when the 
pressure is the same (and in particular when it is zero) above the 
mercury in the two portions of the apparatus, which then form two 
communicating vessels for the mercury; in that case, apart from 
other corrections, the mercury menisci must stand at the same height. 
Another control consists of determining the barometric height which 
is read from Har (Pl. D by producing a vacuum above the mercury 
in the volumenometer. These controls allow one to judge the accu- 
racy of the data that are necessary for the calculation of the cor- 
rections. The uses to which the clips /,, /,,/,,/; and the taps &,, k., 
he, k,,/, and /, (during part of the observations the latter two were 
replaced by clips) were put during the volumetrie operations and 
particularly during the control measurements just mentioned need 
no further description. We may also mention that two pieces of 
wood were used to compress the rubber tube by being screwed 
closer together and thus cause the mercury to rise a little so that 
(following Rayrrien’s method) menisci that were easily readable 
could be obtained in the desired position: the clips themselves could 
also be used for the same purpose. For further details regarding 
the volumenometer and its working we may refer to the papers 
already mentioned and also to the forthcoming Communication an- 
nounced in $ 3. 
The volumenometer has twice been calibrated below the zero mark, 
once for the experiments published in Comms. Nos. 92 and 88, and 
again for the experiments by KAMERLINGH ONNBs and pr Haas on 
the compressibility of hydrogen vapour. The difference between these 
two calibrations was less than '/,,,,,. In this we have allowed for 
the fact that the temperature of the upper portion of the volumeno- 
meter (above the level §) differs from that obtaining in the bath 
surrounding /#. For the purpose of the compressibility research the 
volume of this upper part has undergone some modification, and we 
have ourselves calibrated the volumes x and «’ above the zero mark. 
We have also used the volumenometer itself to calibrate the dead spaces 
kek, k,—k,—k,, and in this, amongst other things, we allowed for 
the number of times that a tap has been turned on being opened. 
By the general application of the method we have given of allowing 
only a low pressure in these spaces a knowledge of their volumes 
with the accuracy that is attainable by adopting these measures is 
necessary only in exceptional cases. 
The pressure of the gas whose volume is fixed by bringing the 
mercury meniscus in the volumenometer to one of the marks made 
