of the Atmosphere. 427 
downwards, just as the direction of the wind in azimuth is fre- 
quently observed to slightly oscillate about its mean position, 
We may conclude, therefore, from sect. 3, that the absolute force 
of the wind is always a little greater than its horizontal intensity, 
as exhibited by the anemometers. 
While such an undulatory motion of the atmospherical eurrents 
may be generally due to the elasticity of the air and the mechani- 
eal influence of terrestrial irregularities, many of my observations 
were such as to clearly show the existence of true upward and 
downward currents. In no other way can we account for ‘the 
steady inclination of the flanges of the anemoscope at times 
when scarcely any horizontal wind was perceptible.. When true 
upward currents were prevalent, the temperature of the air 
was usually increasing and the weather fine. Downward cur- 
rents seemed to be usually preceded or accompanied by a sudden 
decrease of temperature, and these currents themselves usually 
receded rain or unfavourable weather. Regular alternations of 
both classes of currents were usual about noon or the forenoon 
of clear days. The explanation of the last circumstance is ex- 
tremely simple. It depends upon the manner in which the at- 
mosphere acquires the greatest part of its heat during the day. 
A small portion of the solar heat is immediately absorbed in 
passing through the air, but the greater part reaches the ground, 
whence it is imparted to the atmosphere immediately touching 
it. The air so heated expands, and consequently, from its re- 
duced density, it tends to penetrate upwards in currents through 
the overlying strata, which at the same time fall downwards to 
fill up the vacancies. A species of convection, analogous to that 
seen in a boiling or heated mass of liquid, is thus developed in 
the air. The trembling of the air, often noticed over steam- 
boilers, close to the chimneys of steam-vessels, and even on walls 
and gravelled walks heated by the mid-day sun, is undoubtedly 
due to the same minute and rapid currents which take part in 
this process of aérial convection. 
6. That there are more important vertical currents engaged 
in promoting exchanges between the upper and lower strata of 
the atmosphere, within a short distance from the earth, appears 
manifest from experiments made by me in May 1858*. ‘Ther- 
mometers were suspended at different heights, and under different 
circumstances of exposure to the supposed currents. On days 
when the sky was clear, and when, consequently, the direct in- 
fluence of the sun in heating the ground was most decided, ob- 
servations were made every minute, and sometimes every half 
minute, during short intervals. More or less rapid oscillations 
* Report of the British Association for 1858. Transactions of Sections, 
p. 36, 
