222 Proceedings of the Royal Society 
for many reasons, must necessarily derange the observations. As 
no such direct observations can be made at all, results can only 
be arrived at from those which are not absolutely definite and con- 
clusive. Aided by theory, some of these may be here suggested. 
Observations were made by Captain James in his cottage at 
Granton during the prevalence of strong winds.* When this took 
place in the form of violent gusts, the instrument placed in one of 
the rooms, where, of course, it was so far confined, showed then a 
considerable lowering of pressure, while that placed outside re- 
mained comparatively unaltered. Could this cottage have been so 
arranged or constructed as to move forward at a uniform rate of 
speed with the gusts, pressure inside would not then at all be 
reduced, because no local confinement or retardation of the air 
would then take place ; its amount would then be the same as that 
shown by the instrument outside. 
Experiments were made by the writer at Craigleith Quarry, near 
Edinburgh, with a barometer placed at the top and at the bottom : 
the depth of the quarry, as compared with its width, is considerable. 
The remarkable difference at these spots was this, that while the 
height of the mercury, as observed at the upper part of the quarry, 
remained uniform, or nearly so, during strong gusts ; at the bottom, 
its oscillations were very great, with much occasional diminution of 
pressure, which then took place when the gusts passed rapidly over 
the upper surface, to which the air was drawn up from the bottom ; 
“ lifting” and accumulation aloft are thus exemplified. At the bottom, 
where the air was confined, retardation took place. Had there been 
no such confinement — had supply there been equally copious as on 
the surface, no such oscillations would have taken place. This is 
also illustrated by a waterspout, as shown in the first paper on this 
subject in 1875. 
Observations to throw light upon the subject in question may be 
made in the following manner : When a strong wind blows over the 
sea, let there be placed at anchor in the direction in which it blows 
a line of vessels with instruments placed on deck to show pressure. 
The wind passing over them may then be represented as moving in 
inclined columns, in which case the barometer does not indicate the 
real mass of the air aloft. Let another vessel now move forward in 
* See “Trans. Roy. Soc. Edin.” vol. xx. p. 377. 
