244 DR. A. A. RAMBAUT ON UNDERGROUND TEMPERATURE AT OXFORD 
During the observations of 1898, the temperature of the box ranged from 15 c *81 to 
1G°*05, whilst in 1899 it stood at 10°*36. The close agreement of these two results, 
therefore, taken at such an interval of time and at temperatures differing so consider¬ 
ably—by about one-sixth of the whole range with which we are concerned—affords a 
further confirmation of the accuracy of the adopted value of the temperature 
coefficient and the general consistency of the apparatus. 
It is clear that in a series of observations, such as we are at present considering, in 
which the highest temperature does not exceed 25° C,, any change in the zero point 
will have considerably more influence on our results than a corresponding variation 
in the boiling point. Also the researches of Professor Callendar, Mr. Griffiths, 
Messrs. Heycock and Neville, and others, show that the change in the fundamental 
interval, even when the instrument is exposed to enormous variations of temperature, 
1 s very slight. 
Our experience with a sixth thermometer (A), which is kept in the observing room, 
points in the same direction. In this instrument the fundamental interval remained 
practically unaltered after the lapse of a year, the actual values found for it being 
In 1898 ..... 10P067 
and in 1899 . 101'059 
On substituting the values found above for E 0 and E : in formula ( a ) we obtain the 
following expressions for the separate thermometers, giving the temperature on the 
platinum scale corresponding to any reading It :— 
Thermometer 1. pt = (P - 0'31)/1 *0098 = E - (0*0097 E + 0*31) 
2. pt = (E - 0*43)/l*0104 = E — (0*0103 E -f 0*43) 
3. pt = (E - 0*49)/l*0111 = E - (0*0110 E -f- 0*48) 
4. pt = (E - 0*33)/l*0105 = E - (0*0104 E + 0*33) 
5. pt = (E - 0*24)/l*0094 = E — (0*0093 E + 0*24) 
The expressions in brackets on the right-hand side are the corrections which must 
be applied to E to obtain the temperature on the platinum scale. They may be 
tabulated in a very simple form for each thermometer, so that the platinum tem¬ 
perature can be at once deduced from the reading of the resistance. 
Table IV.— Seduction to Platinum Scale. 
Thermometer . 
1. 
O 
3. 
4. 
5. 
R = -10 . . 
-0-21 
-0-33 
-0-37 
-0-23 
-0-15 
0 . . 
-0-31 
-0-43 
-0-48 
-0-33 
-0-24 
+ 10 . . 
- 0 - 41 
- 0-53 
-0-59 
-0-43 
-0-33 
+ 20 . . 
-0-50 
-0-64 
- 0-70 
-0-54 
-0-43 
+ 30 . . 
-0-60 
-0-74 
-0-81 
-0-64 
-0-52 
