356  Kohlrausch and Loomis—Influence of Temperature 
temperatures. Nearly the same conditions prevailed, however, 
during all the series of observations. The temperature of the 
room differed from the mean at most 5°, and consequently 
ap- 
ature, 
Tn like manner, for the calculation of the third correction in 
the formula (p. 355), it was permissible to take t,=20° as a 
mean for all the observations. 
The corrections may now be very much simplified by uniting 
all =e in tabular form, viz: for reading, 
— ee eS ew 0 
the correction is 
O°-0F 04 <0°U 11> -1°5 1° 98 aR O-B%1  -3°°8 
The interpolation for intermediate temperatures was accom- 
plished by means of a graphical representation of this table. 
In one case, however, when the apparatus was filled with water, 
and the velocity of refrigeration could be neglected, the cor- 
rections 1 and 3 were applied separately. 
Correction of the observed periods of vibration.—1. In the pres- 
ent investigations we are seeking the variations of the modulus 
of elasticity caused by a change of eee whereas we 
observe the directive force of the whole wire, which, in eee 
quence of the heating, experiences a dilatation, as well as 
change in its elasticity. ° Since this directive force of the ies 
depends on the length and section as well as on the modulus of 
elasticity, it is necessary to investigate what corrections are 
thereby necessitated. 
It is an interesting fact that no correction at all is necessary. 
The dilatation in length and breadth completely neutralize 
each other. 
If we denote by 
I the ig of the wire, 
r the 8, 
m the ik! tof the unity of le 
K the moment of inertia of the vibrating weight, 
¢ the time of vibration, 
*Tt is merely an accidental coincidence that there is no correction to be ap- 
i to the reading 0°. The mean of the zero points of the three Gusraiuactert 
happens to be perfectly Peeing 
