ANEROID BAROMETER IN GEOLOGICAL SURVEYING. 129 
the outer shows corresponding differences of elevation in feet ; 
for greater accuracy in reading, a vernier is usually attached. 
Instruments of this kind are to be had, which will plainly indicate 
differences of one or two feet in elevation. It is of course 
understood that these readings represent relative elevations only. 
Equipped with an accurate land-survey map and an aneroid, 
the detailing of the topography of an area would be a simple 
matter, if it were not for the numerous sources of error which 
inhere in all barometers, but more especially in the aneroids. 
These sources of error may be classified as mechanical, observa- 
tional, thermal, and atmospheric. They will be discussed in 
order, and the means of controlling them indicated. 
Mechancal sources of error — However perfect the mechanical 
construction of an aneroid may be, a slight shifting or cramping 
of its parts may give rise to an amount of friction which will 
prevent the registering of small changes in pressure, and when 
the change becomes sufficiently great to overcome the friction, 
the index will jump to its proper place. This difficulty may be 
overcome by gently tapping the face and side of the instrument 
before eaeh reading. It is to be borne in mind that while a 
gentle tap will release the mechanism, a more vigorous one is 
very liable to produce the error we are seeking to avoid. 
Another mechanical source of error arises from slight changes 
which are continually taking place in the elasticity of the dia- 
phragms. To avoid this the aneroid should be tested from time 
to time on known elevations, and the tension of the diaphragms 
regulated by a small screw at the back of the instrument, until 
the elevation is correctly given. 
Errors in observation—TVhe weight of the levers is in part 
borne by the diaphragms, and must cause movement in them 
when the position of the instrument is shifted, hence all readings 
should be taken with the instrument in the same position, prefer- 
ably with the dial horizontal. This isa matter of great impor- 
tance, for the reading of a delicate instrument may be changed 
one hundred feet by simply reversing its position. If, however, 
the instrument is uniformly read from the same position, the 
