82 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
This agrees with what has been observed by Dr. Grunmach of 
the Berlin Aichungs Commission. He found the maximum differ- 
ence on a certain thermometer to be + 0.°12 at 29.°8, and another 
secondary maximum of — 0.°04 at 82°. 
A slight change in the values of « and y will make a large 
change in the position of the secondary maximum, For another 
thermometer investigated at the Signal Office this point was found 
to be at 130°. 
The values of « and y can be analyzed still further to ascertain 
whether consistent with known physical properties of glass and 
mercury. | 
Taking Broch’s values for the expansion of mercury, which are 
based on a re-reduction of Regnault’s observations, it is found, 
adapting the figures to the notation used here, that: 
y, = + 0.000181792 
y, = + 0.000000000175 
yg = + 0.000000000035116 
Substituting these values of y, and 7, in the equations 
and aking 3 = 0.000026 we get 
8, = + 0.000000021859 
2, = + 0.000000000099512 
The linear coefficient of expansion of a specimen of glass, such 
as is used in barometer-tubes and thermometers, has been very care- 
fully determined by Dr. Benoit of the International Bureau of 
Weights and Measures. Deriving from this the cubical coefficient,, 
in the notation used here, we have Dr. Benoit’s values 
2, = -+ 0.0000252 
8, = + 0.0000000144 
It will thus be seen that there is a good agreement between the 
two values of , found by the two different processes. 
Remarks were made by Messrs. Gippert, Harkness, and 
PAUL. 
