221 
of Edinburgh , Session 1878 - 79 . 
reaching a point 187 miles north of Christiania, and nearly 1000 
miles from Paris in a direct line, the balloon must have sailed at the 
rate of at least 66 miles an hour, while nowhere did the extreme 
speed of the wind registered at any observatory at all approach the 
speed of the balloon. When the stratum of clouds overhead is of 
sufficient depth, and when the difference in the velocity of their 
different strata is sufficiently great, a columnar inclination of the 
clouds in the direction in which they move will be observed. This 
will illustrate the inclination of the columns of air which repre- 
sent the increasing upward velocity of the winds, and is so often 
referred to. 
Over the ocean, as seamen are well aware, sea birds, when strong 
winds prevail, attempt to fly only on the surface, where their 
velocity is retarded, while above this they are unable to move 
against them. 
To ascertain the amount of diminished or fictitious pressure 
which is due to lifting, the real weight of the same mass of atmo- 
sphere must be ascertained both when it is at rest and when it is in 
motion. The results of such an observation, if made by a series of 
barometers placed vertically above each other to a great height, and 
not very far apart, will, in each case, exhibit an equal amount of 
pressure, and when the atmosphere is at rest, the barometer on the 
surface will then alone show its true weight, but when it is in motion, 
the barometer on the surface will then indicate diminished or ficti- 
tious pressure, because the air is then so far lifted and rarefied 
immediately above the surface, and is accumulated aloft with 
increase of pressure there. The result of this, as shown, is an 
upward abnormal diminution of pressure. The barometer always 
shows the real pressure of the air in contact with its cistern, which 
it does in this latter instance, but it does not here show the weight 
of the superincumbent air, while in the former case it showed both 
accurately. In this way, and also for other reasons, it does not 
indicate correctly the height of mountains, nor can observations 
made at some height and then reduced to sea-level be at all depend- 
able. 
Observations with a series of barometers as thus just above sug- 
gested, will be best carried out in a wide open atmosphere over the 
sea, and not in contact with the slopes of a lofty mountain, which, 
