292 
The following table gives the results of the experiment : 
P2 8, Ry Rg A | B A—B P, 0, y 
74.14 294.— 2 100100 94.6 = 69.6 78.44 289.4 1.3888 
75.26 294.05 yee 100100 95.9 24. 71.9 79.70 289.35 1.3910 
74.72 294.2 ail 109500 89.5 30.- 59.5 78.39 290.34 1.3803 
74.5 293.9 Pit 103800 89.7 30.- 59.7 78.19 289.81 1.4802 
75.19 293.95 Pia 103800 90.9 29.2 61.7 79.— 289.71 1.4160 
75.49 293.1 3.2 100300 145.- 46.4 98.6 81.58 286.6 1.4042 
74.99 293.7 Bag 100500 145.- 47.- 98.- 81.04 287.47 1.3872 
74.99 293.5 AD. 100500 144.7 46.3 98.4 81.07 28701 1.3945 
75.06 293.6 4.7 101200 169.- Dat 146.9 84.14 284.3 1.3927 
74.84 294.1 4.7 100700 170.- Des 148.2 84.00 284.51 1.4027 
75.52 293.8 4.- 190000 158.1 33.6 124.5 83.21 286.2 1.3704 
74.76 ag 5. i 174.2 18.2 156.2 84.41 283.89 | 1.3938 
74.76 4.- te 159.0 33.5 124.5 82.45 285.81 1.3921 
74.76 “(* 3.- 4 142.0 50.- 92.- 80.44 287.81 1.3911 
74.71 ss 6.- 195.- pe 188.- 86.33 281.96 1.3974 
74.71 ay 6.- a 195.- 7.- 188.- 86.33 281.88 1.4013 
74.57 “9 5.- sf 178.5 29.3 154.6 84.12 283.86 1.4001 
74.57 i 4.- : 162.1 40.3 121.8 82.09 285.84 1.3997 
74.57 oe 3.- a 146.5 57.3 89.2 80.08 | 287.84 1.4040 
74.57 a 2.- We 132.5 one 58.8 78.20 289.83 1.4005 
74.44 293.82 i= 118.2 88.8 29.8 76.25 | 291.83 1.3925 
AVGCIrage@)s....0%.0a08 «23. cae 1.3957 
Av. for last 10)... 7.5 tesa eee 1.3973 
P, is the reading of the barometer. 0, is the original temperature 
of the bath. A and B readings of the oil manometer. P, =P. (A—B) d,/d, 
where di = density of oil and d, = density of mercury. 0, is the tempera- 
ture as given by the thermo-couple required to secure a balance. 
y = ratio of specific heats. 
DISCUSSION OF RESULTS. 
The limits of precision seem to be the precision of the resistance, the 
precision of the temperature reading as read by a thermometer, the pre- 
cision of the temperature readings used to calibrate the couple, the con- 
sistency of the I. M. F. of the battery and the density of oil. 
The precision of the resistance was none too good. ‘The last ten ob- 
servations were made with the best box available and R, kept constant so 
that errors due to R. were obviated. The maximum variation of the mean 
of these observations is less than 4 per cent.. and shows remarkable con- 
stancy. 
