37 
the time occupied in calculation on each occasion, he had constructed 
tables from which he could, with very little trouble, read off the corrected 
result. Abstracts of these tables were exhibited (vid. p 38). It was his habit 
to make at least two independent observations, and a difference of ten feet 
in the two results would be sufficient to cause him to go over the ground 
again ; frequently they agreed within two or three feet. 
Mr. Sanders then proceeded to describe the construction of the Aneroid 
Barometer, illustrating his remarks with an instrument kindly lent by 
Messrs. Husbands and Clarke. It consisted essentially of a flat cylindrical 
box from which the air was almost completely exhausted. The variations 
in the atmospheric pressure caused one side of the box to bulge in or out, 
and the amount of this movement was measured by some very exact 
mechanism, much finer than watchwork, which magnified greatly an 
almost infinitely small motion of the side of the box. It was not neces- 
sary to hang the instruments perpendicularly, but always to read them in 
one position, and it was desirable also to free the mechanism by gently 
tapping them. The index on the face occupied O'Ol inch of space, but 
with proper precautions, the indications could be read to 0'005 inch, and, 
with care, good work could be done with Aneroids, though they were occa- 
sionally capricious, and unreliable. 
Where an absolutely correct result was not required, the following 
formula for the Aneroid was found to give a near approximation to the 
truth. 
If S ===' sum of the heights, 
D = difference of the heights, then 
8 : D : : 55000 : a' (height required) 
and, where a temperature correction was to be made, 
x (t — 55°) = corrected height, 
where t == mean temperature. 
This correction was positive or negative, according as t was greater or 
less than 55°. 
Mr. Sanders concluded his communication by recommending Belleville's 
Manual of the Mercurial and Aneroid Barometers, as containing much 
good information' on the subject. 
