Vol. 8, 1922 PHYSICS: C. BARUS 15 
(5) y = 2TrlR''{l-{-s-t/r)Ay/LMT^ 
showing that I virtually enters in the first power only. 
3. Observations. — Large attracting masses M cooperating. The constants 
entering equation (1) have been given for the needle. The new data are 
M = 3368 grams R = 5.m cm. 
Hence the following relations result: 
^' = 10-8 X 4.960 A:y'; and t/r^ .0241, 
which in equation (3) gives 
^ = 10-8 X 4.844 Ay\ 
The second mass was a little smaller; viz., M'' = 2947 grams and R = 5.4:3 
cm. 
Hence the approximate value of K" is 10"^ 5 jgg^ Xhe value of 
t/r is .0236 so that in view of the factor 1/1.0236 
y = K''Ay'' = 10-^ X 5.078Ay'\ 
If the baUs, M, act together 
7 = Ay(l/K' + = 10-8 X 2.479 Ay 
which is partically the fourth part oi K' K". The normal deflection 
should be 7/i^ = 2.69 cm. 
The observations, begun on July 7 and continued for some time, 3 
times daily, forenoon (A), afternoon (P) and night (A/"), with periods of 
from 30 minutes to one hour between consecutive observations,^ are 
given in the figure, the mean value of Ay being inscribed on each group of 
observations terminating in little circles. As the normal value should be 
A;v = 2.69 cm., it is obvious that only the night (A^ observations are ac- 
ceptable. On July 7 the apparatus, which had just been set up, was slowly 
approaching thermal equilibrium. 
If we exclude the first night datum as the apparatus was too fresh and 
the exceptional datum during the fog of July 9, and take the last doublets 
specified, the mean of the eight deflections is A3; = 2.666 ±.0017 cm. 
Therefore 
-y = 10-8 X 2.479 A3/= 10-8 X 6.609 
with a mean error in T and Ay together, of less than .3%. This is 
as near the normal value as one may hope to get, seeing that the small 
individual deflections, Ay, must be read off on a millimeter scale. If all 
night observations were taken 7 = 6.63 would result. 
Night observations made on an exhausted gravitational apparatus, 
therefore, give promise of trustworthy results, particularly if a season of 
uniform temperature is chosen and a relatively finer quartz fibre is in- 
serted. While the exhaustion here in question has not been carried 
further than a few millimeters, the experiments are all aiming at the 
final test under complete exhaustion. 
* Advance note from a report to the Carnegie Institution of Washington. 
t To secure better vision the scale was slightly shifted on July 13. 
