291 
and the temperature of the bath changed until another balance was found. 
In this way a number of balances were obtained for different values of Ri. 
By plotting Ri: against the difference in temperature of the two junctions a 
curve was obtained which gave the temperature for any resistance. The 
calibration was made with a number of different couples and the results 
were entirely consistent, no point being off the straight line thus found 
more than 1-20 of a degree centigrade. 
The pressure in the bottle was measured by means of an oil manom- 
eter; as considerable time was consumed by the oil coming to a steady 
state, it was deemed desirable to place a stopcock between the manometer 
and the bottle, and after the pressure was determined cut off the mano- 
meter before expansion. The pressure for the following trial was ad- 
justed approximately by an auxiliary mercurial manometer and the final 
adjustment was made with oil. The use of the oil manometer was neces- 
sary, as the errors introduced in the reading of the mercurial manometer 
were of a higher degree of magnitude than was permissible. 
The delicacy of the apparatus was indicated by the fact that the 
observer could readily detect a difference in pressure of 2 mm. of oil, 
density .84. 
The value of y was determined as follows:— 
From the adiabatic law, P Vy = a constant 
From the law of Charles, PV = RT 
Ore iV 176—=)PeVa} 
Wioi\oy 1B 
then { — } =— 
Vi Pe 
bute. 2nVin— Ray 
and PeVe= RT2 
Ve Biase 
Vi ile Jegy 0'n 
Ve\}y (— SIUING Vee ext 
Vi * 124) Abi flaps 
Pi 
log — 
2 
therefore y = 
12 Ti 
log — — log — 
Pe Te 
