( 1096 ) 
to drive the vapour up between the coils of the regenerator spiral 
was found, from measurements made with air, to be less than */,, 
mm., and may, therefore, be left out of account. Only at very low 
pressures will a correction be necessary for it in obtaining very 
accurate results. 
Fig. 2, Pl. II shows the apparatus that serves to regulate the 
quantity of gas pumped off through the exit valves; by such regu- 
lation the temperature of the cryostat part of the apparatus is kept 
as coustant as possible, and from its constancy one can judge how 
far the vapour pressure remains invariable; it also shows how this 
pressure is measured. For pressures greater than 5 em. the gauge /, 
is used, for pressures between 5 em. and 1 em. /,, and for pressures 
lower than 1 ‘em: \7,.: By opening Ki, Wiens Kin Or AGS, ae 
Kiss, Joa OV Iya, as the case may be, is brought to a definite pressure, 
which is measured by J, or by the Mac Leop gauge /,; K;,, or Ki, 
is then closed, and the taps regulating the rate at which the gas is 
pumped off are operated so that the oil in the graduated sloping 
tubes of the indicators remains at the same mark. 
Any definite pressure and, therefore, any definite temperature at 
the surface of the liquid in the cryostat can be quite satisfactorily 
obtained. The temperature of the bath, however, is less assured, 
since stirring is not possible, and the conductivity of Cu offers 
but slight compensation for this defect. The lower parts of the 
bath are at a higher temperature, and the corresponding vapour 
pressure may be increased by 0.009 to 0.011 mm. of mercury 
per mm. ‘distance from the surface of the liquid helium; at the 
lowest temperatures this is equivalent to a temperature difference 
of 0.06 degree, and with this uncertainty we must be content as 
long as we have not at our disposal a cryostat in which stirring is 
possible; it is not, however, greater than uncertainties arising from 
other causes that are already present. 
§ 2. The Thermometer. Temperatures were measured by means 
of a constant volume helium thermometer of zero pressure = 14.5 
em. (Cf. Comm. N°. 112). 
At the lowest measured temperature, the pressure of the gas in 
the thermometer was 1.2 mm., and the vapour pressure of the 
helium was only 2 mm. The circumstances of measurement were, 
therefore, pretty much the same with respect to helium as if a con- 
stant volume ether vapour thermometer were used for determining 
ordinary atmospheric temperatures. In the present case there is, 
moreover, the particularly small value of the pressure itself to be 
