of Edinburgh, Session 1878 - 79 . 
223 
the direction and at the same rate of speed as that of the wind, a 
calm will now prevail over its deck, and the columns above it may 
then be regarded as being vertical, and indicative of the real weight 
of the air aloft ; when it passes close alongside of those at anchor, a 
difference of pressure will be found on each, although the mass of 
air aloft is the same. For various reasons, however, the real amount 
of difference here will not in this way be correctly ascertained, and 
it will not be great. If a calm prevails over the vessels at anchor 
they will then show the real weight aloft. The moving vessel will 
now produce and encounter a force of wind equal to her rate of 
speed, and will not now show real weight aloft. 
Vessels crossing the Atlantic against the strong prevailing west 
winds, will increase their force in proportion to the rate at which 
they sail against them. They must, therefore, indicate a difference 
of pressure when passing those vessels returning with the winds in 
their favour. Its amount will not be great, because the extent of 
the resisting surface which is represented by their decks is insuffi- 
cient to produce retardation of horizontal and lateral supply. If, 
however, the instruments are placed in some enclosure, as was seen 
in the cottage, some amount of difference of pressure will then be 
shown. 
For want of space here, and owing to the extreme difficulty of 
the subject, it cannot here be fully explained, but it may be stated 
that the conclusion has been arrived at, that as a general rule, 
the barometer in the Tropics, with, of course, exceptional cases, 
more nearly shows the real weight of the atmosphere than it does 
in more northerly latitudes. In countries of very different descrip- 
tions of structure and climate, this will probably be found to be the 
case, and also over the land and the sea. 
The isobars, which are constructed on the supposition that the 
barometer always shows the real weight of the air, cannot be accu- 
rate. In the different segments which surround an area of low 
pressure, they will require a difference in the amount of their cor- 
rection.* In the same way, and for other reasons also, gradients 
which are figuratively equal, as is well known, do not exhibit the 
same amount of incline, or the same amount of inflow of the winds. 
Correction here is evidently also necessary . — May 1878. 
* See “ Proc. Roy. Soc. Edin.” 1874-75, p. 618. 
2 B 
VOL X. 
