DETERMINATION OF HEIGHTS BY BAROMETRIC METHODS. 703 
and consequently they neutralise each other. The writer’s 
experience does not support this view completely. The meniscus 
at either end is formed under different circumstances; if the 
mercury has been rising in the one it has necessarily been falling 
in the other, and examination will at once show that the shape of 
the meniscus, and presumably the amount of the depression, 
differs considerably under these circumstances. The writer finds 
that siphon mountain mercurials (which, to be portable, are neces- 
sarily of small bore) require vigorous tapping. Further, when 
the instrument is inverted and ready for travelling the (normally) 
lower limb is free from mercury, whilst, when in use, the mercury 
will run up that limb to an extent depending upon the atmospheric 
pressure. Hence this portion of the mercury is brought succes- 
sively in contact with a very large area of tube in proportion to 
its mass, and it speedily becomes more or less foul when in use. 
The advantage possessed by the siphon over the Fortin mercurial 
for mountain work, in the writer’s opinion, is that owing to the 
absence of the cistern at the bottom the volume of mercury 
required and consequently the weight of the instrument is very 
much reduced. Latterly the writer has been using a Fortin 
mountain mercurial, in addition to the siphon ones, and so far as 
his experience goes he decidedly prefers it. In mounting the 
instruments prior to reading it is of course essential to have 
them perfectly vertical, and, in the writer’s opinion, made 
fast at both top and bottom. Sufficient steadimess cannot be 
obtained by supporting the instrument in gimbals, as is frequently 
done, or by suspending them from the top only. The writer's 
system of plumbing a Fortin barometer is as follows :—Having 
suspended the instrument and made it fast at the bottom, set the 
mercury to touch the fiducial point, then rotate the instrument 
through 180 degrees on its vertical axis. If the setting is still 
perfect the instrument is perpendicular in the vertical plane 
passing through the centre of the tube and the fiducial 
point. If not, alter the bottom clamping screws until the 
above adjustment remains correct on rotation through 180 
degrees. Then rotate through 90 degrees and adjust in plane 
perpendicular to the first. Care must be taken in erecting the 
instruments to place them so that the sun will not shine on 
them when in use, as otherwise the temperature of the mercury 
in the tube will probably differ from that of the attached thermo- 
meter. It is scarcely necessary to say that they must always be 
inverted prior to being carried. If a siphon barometer it should 
always be inverted or restored to the normal position with the 
short hmb wpwards, otherwise air is certain to get in the tube. 
If a Fortin, the mercury should be caused to fill the cistern (by 
using the adjusting-screw) prior to inversion. Mercurial baro- 
meters require most careful carrying. When carried in a vehicle 
