1913-14.] Principia Atmospherica. 83 
can be said is that there is no example of the approach to such an ac- 
cumulation. There are a sufficient number of examples in which there is 
a reversal of fall of temperature just below the stratosphere, and these 
show that the stratosphere has, if anything, a little to spare in the way of 
resistance against penetration. Hence, from the point of view of meteoro- 
logical theory we regard the stratosphere as impenetrable. 
5. The Law of Saturation. 
The amount of water vapour contained in a given volume of air 
cannot exceed a certain limit which depends upon the temperature and 
upon nothing else. 
This is really simply a statement of Dalton’s law of the saturation of a 
gas with the vapour of a liquid, but it is quoted here partly because it 
refers to the only form of variation in chemical composition to which the 
meteorological atmosphere is subject, and also partly in order to avoid a 
misapprehension that is very widespread. It is a well-known physical 
principle that when a vapour is condensed the “ latent heat of vaporisation,” 
which, in the case of water vapour, is very large, is liberated. The state- 
ment of the principle is not complete ; it should go on to say that the 
condensation cannot take place unless provision has been made for dispos- 
ing of the heat which will be liberated. In the case of the atmosphere it 
is often assumed that no provision of the kind is required, and that the air 
will, in consequence, be warmed by the heat set free. Herein lies the mis- 
apprehension. Vapour of water in air will not condense unless the air is 
cooled, and the amount of condensation will be limited by the amount of 
the cooling. 
It should, however, be noted that the wording of the law as here given, 
namely, that the limiting amount of water vapour depends upon the 
temperature and upon nothing else, implies a statement about the atmo- 
sphere about which it is necessary to be explicit. Since Dalton’s law was 
enunciated, the researches of Aitken and others have shown that the cooling 
of a mass of air below the “ saturation point ” causes condensation only if 
there are nuclei upon which drops of water can form. In the absence of 
such nuclei, laboratory experiments have shown that condensation does not 
take place until the limits of saturation have been largely exceeded ; “ four- 
fold saturation ” is necessary in such a case. Air without nuclei cooled 
below its “ saturation point ” is said to be supersaturated, and the statement 
of the law of saturation as set out implies that supersaturation does not 
exist in the free air. This is another case in which there is no physical 
reason to prevent anyone imagining circumstances in which supersaturation 
