AT NIGHT, FROM THE EARTH, ETC. 
123 
By this means the comparisons were made with ease and certainty, and they readily 
indicated every inequality of the thermometer tubes. 
1843. 
Greenwich mean time Oct. 21*. 
Correction to be added to read- 
ings of the thermometers. 
Name or letter 
of reference of 
the thermometer. 
7 h 30 m . 
ll h 30 m . 
15 h 30 m . 
17“ 40 m 
Standard. 
45-1 
49-8 
69-3 
90-5 
A. 
45-1 
49-8 
69*3 
90-5 
B. 
45-1 
49-8 
69-3 
90-5 
C. 
45-1 
49-8 
69*2 
90-3 
D. 
45-0 
49-8 
69-3 
90-4 
E. 
45-1 
49-8 
69-3 
90-3 
F. 
45-1 
49-8 
69-3 
90-4 
G. 
43-0 
49-8 
69-3 
90-4 
H. 
45-1 
49-9 
69-3 
90-3 
I. 
45-1 
49-8 
69-2 
90-4 
J. 
45-0 
49-7 
69-3 
90*4 
K. 
45-0 
49-7 
69-3 
90-4 
L. 
45-0 
49-8 
69-3 
90-4 
M. 
45-0 
49-8 
69-3 
90-4 
N. 
45-1 
49-8 
69-3 
90-5 
0. 
45-1 
49-9 
69-4 
90-5 
P. 
44-6 
49-4 
69-0 
90-0 
Add 0 o, 5 to all readings. 
Q. 
44-9 
49-6 
69-2 
90-3 
Add 0 O- 2 to all readings. 
R. 
45-0 
49-8 
69-3 
90-5 
S. 
44-8 
49-5 
69-0 
90-2 
Add 0 o, 3 to all readings. 
T. 
44*8 
49-5 
69-0 
90-2 
Add 0 o, 3 to all readings. 
U. 
45-1 
49-8 
69-2 
90-5 
V. 
45-1 
49-6 
69-2 
90-5 
vv. 
43-1 
49-7 
69-1 
90-5 
X. 
45-1 
49-8 
69-1 
90-5 
Y. 
45-1 
49-6 
69-0 
90-4 
During- the operation of comparing the thermometers, it was found absolutely 
necessary to have all of them in the same position with respect to the vertical, for it 
sometimes happened that their readings varied with their position ; this variation 
with the self-registering minimum thermometers was so great that I could not com- 
pare them at this time ; afterwards I procured a shallow vessel of sufficient extent 
to receive them horizontally ; in this vessel the comparisons were taken, and their 
readings were found to agree with those of the standard to 0°T or 0 o, 2, the same 
thermometer being frequently as much in excess at one part of the stem, and as much 
in defect at another. 
The possession of upwards of fifty instruments whose extreme difference of reading 
from the standard was a constant quantity of half a degree in one thermometer, and 
a constant quantity of 0 o, 2 or of 0 o, 3 in three others, the remainder being absolutely 
free from error, was exceedingly gratifying. I felt that all the time and trouble I 
had bestowed on them was well repaid. I had not expected to be so successful in 
obtaining so many essentially good thermometers, and I had prepared myself to 
ascertain their errors at every division by Bessel’s formula, which, although implying 
a long process, gives good results ; all this trouble, however, I was saved, and also 
