118 
be equal to 25-3916 apparent mm. where 
1 + 16'67a 
a = -0000187) at any common temperature. From the mean 
of the following comparisons it was found that '052 has to 
be subtracted from the reading of the metrical scale to make 
it correct, which agrees well with the correction given in 
the Kew certificate, namely — '05. 
Inches. 
MM. 
(Calculated) 
CD 
MM. 
(Observed mean) 
(2) 
MM. 
Difference 
(2) — (1) 
28 
710-965 
711-006 
+ •041 
28*5 
723-661 
723-720 
+ -059 
29 
736-356 
736-400 
+ -044 
29'5 
749-052 
749-110 
+ •058 
30 
761-748 
761-817 
+ "069 
30'5 
774-444 
774-500 
+ -056 
31 
787-140 
787-180 
+ •040 
+ -052 
The actual comparison of the two barometers. 
The mode of procedure. This was done in the following 
order : — 
A Take reading of the upper surface of mercury of the 
standard barometer. 
B Take reading of the working barometer. 
G Adjust the screw-head of the standard. 
D Repeat A. 
E Repeat B. 
F Note temperatures. 
G Read the height of the centre of cross of the screw- 
head (twice). 
The temperature was observed by the thermometer 
affixed to the working barometer, another plunged in mer- 
cury of the standard, and the third hung in contact with 
the cathetometer scale. The last thermometer was care- 
fully shielded from the direct radiation from the body of 
the observer. 
