Voiv. 8, 1922 
PHYSICS: C. BARUS 
e315 
are constructed in figure 1, on the right, in seconds. The variations 
may run from 6 to 7 minutes and the T/2 values are largest in the morning 
and least at night. They march about in parallel to the X values, if either 
the R or the F position alone is taken. T thus also varies harmonically 
in the lapse of time with a period of 24 hours referable to dO/dt, provided 
the first semi-period, T/2 is taken. 
From these results it might be surmised that the true period T can be 
computed from the log decrement observed at the same time. But it is 
easily seen that the correction so obtained is of an order entirely too small 
to account for the existing divergencies. The theory of the damped 
pendulum is only useful here as an analogy. 
4. Static Displacements, Ay. — Continuous series of these data were 
given in the earlier papers and obtained from night observations (7-11 
P.M.), when they are liable to be most constant. The black dots in figure 
1 are the present values for different parts of the same day. Ay is the 
static displacement finally obtained, when the attracting weights M are 
swung from one side of the needle to the other. As figure 1 shows, Ay 
is now also markedly harmonic in the lapse of time, or of the reduced ex- 
haustion proportional to time, with a period of one day. 
We thus come to the result that X, T/2, Ay are similar time functions; 
that as a first approximation they are independent of the change of vis- 
cosity of rarified air and that their relative variation is such as to admit 
of their expression in terms of each other. It does not necessarily follow 
that a correct value computed for one (for instance for X) would lead to 
correct values for T/2 and Ay; but it is a project well worth testing. 
5. Comparison of X, T/2, Ay. — It is next necessary, therefore, to more 
specifically compare all the quantities obtained in the present investi- 
gation. This has been done in figure 2, where the abscissas are X log e and the 
ordinates T/2 and Ay, respectively. 
