648 
PROFESSOR W. RAMSAY AND DR. J. SHIELDS ON THE 
temperature, t is zero ; hence the origin of the temperature-scale (r) is also the origin 
of the scale of tension. 
But the above equation, unmodified, represents facts with only approximate 
accuracy, for reasons which will now be adduced. 
Let ys and r be the axes of the curve AO, showing the relation between ys and t. 
At O, the critical temperature, t = 0, and the value of ys is also 0. With rise of r, 
^.e., with fall of temperature, ys Increases slowly at fi.rst, but soon attains a nearly 
steady rate of increase, pictured by the nearly straight line CA. The origin of the 
Fig. 1. 
line CA, which may be regarded as a tangent to a curve at some point intermediate 
between A and C, is not 0, but B, a temperature some degrees on the ordinary scale 
below O, the critical temperature. Hence, in representing the slope of the line CA 
by an equation, r cannot be one of the factors, but r diminished by the number of 
degrees between 0 and B. The equation then becomes, where d represents that 
number, 
ys — K {t — d). 
A similar correction might possibly be applicable to the gaseous equation, if we 
wish to exhibit the behaviour of gases at ordinary temperatures, which are far above 
the absolute zero of the thermo-dynamic scale. Were such a correction applicable, it 
would have the effect of merely altering the zero-point on the gas thermometer by 
some degrees. 
That d {ysjdt) is a constant has been shown to be approximately true by Eotvos 
(‘Wied. Ann.,’ 27, 452); he also points out that 5 may be taken as equal to (Mr)L 
where v is the molecular volume of the liquid under investigation, i.e., the volume in 
cub. centims. occupied by molecular weight of the compound taken in grammes. He 
deduces from the constancy of the ditferential the equation y (Mv)^ = k [T — T'), 
where T is the critical temperature of the liquid, and T' the temperature at which 
y and (Mv)^ are observed, i.e. (T — T') = r. But, as we shall show, such an equation 
leads to erroneous results. 
