190 Mr. J. F. Daniell on the 
In endeavouring at the same time to account for these pheno- 
mena upon the known laws of aériform fluids, I was led to con- 
struct an hypothesis which appeared to me to explain this gradual 
decrease of the oscillations ; but at the same time pointed out a 
condition of the problem which would at once, if confirmed by ex- 
perience, be a test of the correctness of the solution. 
Let us suppose that in the atmosphere surrounding the earth a 
circulation is kept up between the poles and the equator; and that 
the cold dense air of the former regions flows in a lower current to 
the latter, while the elastic air of the latter is returned in an upper 
current to the former. There can be no difficulty in imagining 
further that, as long as these currents are maintained with regular 
velocities, a barometer, at all intermediate stations, might exhibit 
an equal pressure of the a€rial columns; for as much air would 
flow from their summits as would be returned to their bases. A 
general alteration of temperature, which equally pervaded both 
currents, would produce no alteration in the weight of a vertical 
section, comprising both; nor would a partial alteration equally 
diffused through the upper and under section of any one column. 
The velocities of the currents would be partially altered thereby, 
but the higher and lower would still compensate each other. But 
an alteration of temperature which affected the upper and lower 
currents unequally, would produce partial expansions and con- 
tractions, which would effect an unequal distribution of the pon- 
derable matter. If the lower stratum of any perpendicular section 
were expanded by heat, while the upper were unaffected, the out- 
going current of that section would be increased, while the in- 
coming current would be checked; and the balance of the two 
being disturbed, the total weight would be diminished. On the 
other hand, a local decrease of temperature would produce the 
analogous contrary effect. Now the alternations of heat and cold, 
produced by the changes of day and night, although they may be 
regarded in a general way as pervading both currents, act with suf- 
ficient inequality to induce us to expect a corresponding fluctuation 
in the weight of the atmosphere at any particular point. The heat- 
ing Surface being below, the warm particles quickly ascend, and 
EE a es 
= ity 
ey 
