56 REPORT—1840. 
begin to fall, and at the moment when the air is parting with its heat 
as fast as it receives it, the barometer will indicate the exact weight of 
the atmosphere. The barometer, however, will continue to descend 
on account of the diminishing tension of the air and consequent sink- 
ing upon itself as the evening advances ; and its greatest depression 
will be at the moment of the most rapid diminution of temperature, 
which will be about 4 or 5 o'clock. 
At this moment the barometer will indicate a less pressure than the 
true weight of the atmosphere. The whole upper parts of the atmo- 
sphere have now acquired a momentum downwards, which will cause 
the barometer to rise above the mean as the motion diminishes, which 
must take place some time in the night. This rise will be small, how- 
ever, compared with that at 9 or 10 am. As the barometer now 
stands above the mean, it must necessarily descend to the mean at the 
moment when it is neither increasing nor diminishing in temperature, 
which will bea little before sunrise. If this is the true explanation of 
the four fluctuations of the barometer in a day, it will follow that the 
morning-rise ought to be greater at considerable elevations, provided 
they are not too great, because some of the air will be lifted above the 
place of observation; and such was found to be the case by Col. 
Sykes in India. 
On the Meteorology of Perth. By Dr. ANDERSON. 
Perth is elevated about 30 feet above the mean level of the ocean, 
and situate in lat. 56° 23' 40" N., long. 3° 26' 20". W. Dr. Anderson 
stated that the magnetic variation, which seemed to have reached its 
maximum in 1815, was 26° 54’ W. in Nov. 1836; and that the mag- 
netic dip was 72° 10' in May 1838. The mean barometrical pressure, 
deduced from a period of six years of consecutive observations, con- 
tinued from 1829 to 1835, was 29°802, the time of observation being 
nine o'clock in the morning. By a comparison of the means of each 
month, for the several years, with the mean of the entire period, it ap- 
peared that the means in defect greatly exceeded those in excess ; from 
which he concluded, that the disturbing causes which produce a dimi- 
nution of atmospherical pressure are more sudden, as well as more 
powerful in their operation, than such as give birth to an opposite con- 
dition,—a result which implies that the causes contributing to a low 
state of the barometer are of limited extent and partial influence,— 
and may be explained by referring them partly to the diminution of 
aerial elasticity occasioned by a rapid condensation of aqueous vapour, 
and partly to the combustion, by electricity, of large portions of car- 
buretted hydrogen in the upper regions of the atmosphere over the 
place of observation. Hence he inferred that an arithmetical mean be- 
tween two observations, the one expressing the highest and the other 
the lowest height of the mercury, will rarely give the true average 
height for the mean interval of time between the observations. 
The mean annual range of the barometer at Perth, derived from a 
period of nine consecutive years, is 2189 inches; but the extreme 
