Temperature Coefficient of the Balance Magnet. 
xliii 
TABLE 14. — Observations for the Temperature Coefficient of the Deflection Magnet. 
Gottingen 
Tempera- 
Reading 
Reading 
of Declino- 
11 " Wq 
Bifilar. 
Mean Time 
ture of 
t 0 ~ t. 
of 
meter 
minus 
1843. 
Magnet. 
Uni filar. 
reduced 
f(d-d 0 ). 
Reading 
Thermo- 
-fd. 
Corrected. 
meter. 
Sc. Div. 
Sc. Div. 
Sc. Div. 
Sc. Dir. 
- 
Nov. 10 23 
24 
Magnet 
away. 
264-39 
157-26 
23 
53 
35-30 
24-65 
16- 90 
17- 90 
10-73 
156-65 
1-80 
1-60 
1-23 
501-0 
42-6 
Nov. 11 0 
0 
13 
30 
59-95 
76-85 
1205 
13-07 
156-17 
155-59 
501- 7 
502- 7 
42-7 
42-8 
0 
46 
58-95 
22-05 
20- 20 
21- 60 
22- 20 
19- 90 
20- 30 
25-05 
21- 05 
24-05 
30-50 
23- 40 
22- 55 
11-88 
155-63 
1-69 
1-48 
1-45 
1-51 
1-29 
1-31 
1-91 
1-21 
1- 61 
2- 21 
1-91 
1-79 
503-3 
42-9 
1 
1 
6 
24 
36-90 
57-10 
10- 05 
11- 68 
155-49 
155-64 
503- 9 
504- 3 
430 
43-2 
1 
42 
78-70 
13-16 
155-67 
505-6 
43-6 
2 
0 
56-50 
11-11 
155-13, 
505-8 
43-8 
2 
18 
36-60 
10-05 
155-36 
506-8 
43-9 
2 
38 
56-90 
11-50 
155-50 
507-4 
44-0 
2 
54 
81-95 
13-41 
155-50 
506-8 
44-3 
3 
14 
60-90 
12-50 
155-80 
507-4 
44-6 
3 
3 
32 
53 
36-85 
67-35 
10-78 
13-30 
155- 69 
156- 00 
507- 7 
508- 0 
44-8 
44-9 
4 
4 
10 
28 
43-95 
66-50 
11-72 
13-88 
156-33 
156-70 
507- 8 
508- 1 
450 
45-0 
4 
45 
Magnet 
away. 
264-35 
157-02 
Correction 
for 1° Fahr. 
0-000287 
•000369 
•000278 
•000309 
•000289 
•000260 
•000272 
•000265 
•000253 
•000307 
•000233 
•000268 
•000287 
•000324 
•000315 
The mean value of q = 0 0002877. 
Balance Magnet. 
61. The value of the coefficient for the balance magnet was obtained by obser- 
vations, similar to those for the deflecting magnet, on 5 different days ; weight was 
given to the result of each day's observations, depending on the formula, No. 32. 
The final result was 
q - 0-000073.* 
62. The impossibility of determining the value of the micrometer divisions of 
the balance magnetometer by the usual method has been already pointed out. It 
is obvious, therefore, that if shut up to this method the temperature corrections 
cannot be applied even when the observations are left in the state of micrometer 
divisions. The doubt whether the changes of the magnetic moment of the needle 
occur as rapidly as those of temperature in all cases, and therefore, whether a co- 
efficient obtained from changes of 30° or 40° in a few minutes (as in the tempera- 
ture experiments) was likely to be applicable to observations where the changing 
temperature of the needle was rarely above 1° in the hour, and the fact that other 
sources of error (perhaps the effect of temperature on the points of support of the 
* Tables 19 and 20, Introduction to the Makerstoun Observations, 1841 and 1842. 
