70 
PROFESSOR HUGH L. CALLENDAR ON 
The following table gives the corrections found for the second half of the sliae-box 
on the second occasion, December 20th, 1894. 
Table III.—Calibration Corrections of Second half of Slide-Box. 
Reading 
of main 
dial. 
Correc¬ 
tion in 
ohms. 
Reading 
of main 
dial. 
Correc¬ 
tion in 
ohms. 
Reading 
of main 
dial. 
Correc¬ 
tion in 
ohms. 
Reading 
of main 
dial. 
Correc¬ 
tion in 
ohms. 
Reading 
of main 
dial. 
Correc¬ 
tion in j 
ohms. 
50 
-3-96 
60 
-4-58 
70 
-3-13 
80 
-1-33 
90 
-0-92 
51 
-4-42 
61 
-4-68 
71 
-3-32 
81 
- 1-21 
91 
-0-96 
52 
-4-57 
62 
-4-07 
72 
-3-19 
82 
-1-11 
92 
-0-45 
53 
-5-09 
63 
-4*04 
73 
-3-03 
83 
- 1-24 
93 
-0-70 
54 
-4-88 
64 
-3-69 
74 
-3-03 
84 
-1-38 
94 
-0-63 
55 
- 5 • 15 
65 
-3-60 
75 
-2-47 
85 
- 1-70 
95 
-0-82 
56 
-5-26 
66 
-3-62 
76 
-2-16 
86 
- 1-83 
96 
-0-94 | 
57 
-5-31 
67 
-3-68 
77 
-1-92 
87 
-1-64 
97 
-0'22 
58 
-5-19 
68 
-3-45 
78 
- 1-61 
88 
- 1-71 
98 
-0-23 
59 
-5-23 
69 
-3-51 
79 
-1-64 
89 
-1-24 
99 
-0-06 
In comparing the differences between the two calibrations given in the last column 
of Table II., it will be noticed that there is a cumulative divergence amounting to 
about half an ohm at the middle of the range. This is the kind of error to be 
expected in this method of calibration. It might be explained by the considerable 
difference in temperature of the box on the two occasions, but it is within the limits 
of error of the first series. The galvanometer was not sufficiently steady on that 
occasion, and the temperature rose nearly half a degree in the course of the 
observations. The observations serve, however, as a satisfactory verification of those 
of Series II. 
It will he observed that the correction does not amount to so much as 1 part 
in 10,000 of the reading at any point of the slide-box, except quite near the 
beginning, a part which was never used in accurate comparisons. Also that the 
change of the correction in passing from one point to the next, never exceeds 
1 part in 100,000 of the reading in the second half of the box, although the 
errors of two or three of the individual 1000-olnn coils exceeded 1 ohm. This is 
due to the levelling effect of the vernier. 
It must be remembered that the corrections were never required beyond the 
nearest ohm, so that a difference of less than ‘5 could be neglected. 
It might be supposed that greater accuracy of calibration would have been 
attained by dividing up the box into subsidiary intervals of 10,000 ohms, and 
comparing each of these intervals with an auxiliary resistance, on the analogy of the 
method usually employed in the calibration of a mercury thermometer. I did not 
find, however, that any advantage was obtained by this procedure, and it is evident, 
on reflection, that the two cases are not precisely analogous. The advantage of 
