81 
191 3-14. J Principia Atmospherica. 
atmosphere is free from explosions and elastic wave-motions, or that their 
effect is so small that it does not enter into meteorological calculation. 
It follows that the numerical values of pressure, temperature, and 
density at any point of the atmosphere are related by the usual formulae 
for the gaseous laws. In other words, when due allowance is made for 
the difference of composition in consequence of the variation in the amount 
of water vapour or other possible causes, the relation p = ROp holds, where 
p, 0 , p are the pressure, temperature (on the absolute scale), and density of 
the air, and R is a “constant” which is altered by an alteration in the 
composition of the air, but not by other causes. 
3. The Law of Convection. 
Convection in the atmosphere is the descent of colder air in con- 
tiguity with air relatively warmer. 
The law is advisedly stated in this form (although objections may be 
taken to it for want of strictness) because the driving power of the 
convective circulation comes from the excess of density of the descending 
portion, and the excess of density in atmospheric air is due in nearly all cases 
to low temperature. Differences of density might be caused by differences 
of pressure or by differences in the amount of moisture contained in equal 
volumes. But finite differences of pressure cannot persist in contiguous 
masses of air ; the amount of water vapour in air at the ordinary tempera- 
tures with which a meteorologist has to deal is only a small fraction of the 
whole mass, and the colder the air is, the less water vapour is required to 
saturate it. Consequently, although it would be possible in a physical 
laboratory to display a sample of air which, though warmer, is yet denser 
than another cooler sample on account of the humidity of the latter, the 
conditions would not easily occur in nature, and the motive power for 
convection would be exceedingly small. Such cases may therefore be left 
out of account, and we may consider that, of two contiguous masses of air, 
the colder is the denser. 
The law of convection is usually stated with regard to the warmer 
part of the convective circulation, and takes the briefer form that warm 
air rises. The general adoption of this briefer form is due to the fact that 
the warming of air at the surface is a matter of common knowledge, and it 
occurs in the daytime, when its effects in producing a local convective 
circulation are often quite distinctly visible. The form which is adopted 
here, however, is preferable, because in any case it is the cooler and heavier 
air in the neighbourhood which must be looked for if the true cause of 
the circulation is to be found ; and, although on the smaller scale the 
VOL. xxxiv. 6 
