Supplement to “ Nature,” April 14, 1923 Vv 

The Water in the Atmosphere.! 
By Dr. G. C. Simpson, F.R.S., Director of the Meteorological Office. 
HE dictionary definition of “ saturation ” is “ the 
state of a body when quite filled with another,” 
and it is usual to think of saturated air as air which is 
full of water vapour to such an extent that further 
water cannot be added without condensation taking 
place. This, however, is a wrong conception, for there 
is no limit to the amount of water vapour which air 
can contain at any temperature, provided that it is 
perfectly pure, except that ultimately the molecules of 
vapour will be so near together that there will be no 
distinction between vapour and liquid. 
Air at 30° C. is said to be saturated when its vapour 
pressure is 31°51 mm. of mercury, but if to such satur- 
ated air we add water in sufficiently small drops it 
will be evaporated immediately and the liquid drops 
transformed into invisible vapour. If we could add 
drops with so small a diameter as 1°6 x 1077 cm. the air_ 
would devour them with avidity until its vapour 
pressure was more than 126 mm., and then would be 
ready for more if smaller drops could be supplied. 
The question of whether air can hold more vapour or 
not depends entirely on how the water is presented to _ 
it. Ifthe water is presented with a flat surface, evapora- 
tion will take place until there is so much water vapour 
present in the air that as many water molecules return 
to the surface as leave it ; the air is then saturated with 
reference to a flat surface of pure water. If the surface 
is curved—convex towards the air—more water mole- 
cules will leave each square centimetre of the surface 
than in the case of a flat surface ; hence there must be 
more vapour molecules in the air before equilibrium is 
attained. Thus the saturation vapour pressure over a 
curved surface is higher than over a flat surface at the 
same temperature. On the other hand, if the water 
contains certain impurities—e.g. sulphuric acid, calcium 
chloride, and salts in general—less molecules leave the 
surface than in the case of pure water ; hence a smaller 
number of molecules in the air will be sufficient to 
produce equilibrium over a solution. 
In accordance with the usual practice we will describe 
air as saturated when the water vapour it contains is in 
equilibrium with a flat surface of pure water at the 
same temperature. This will define the saturation 
pressure at each temperature, and relative humidities 
will be given in terms of this saturation pressure. 
It is well known that water can exist in the liquid 
state at temperatures far below the freezing point, and 
therefore water and ice may exist side by side over a 
large range of temperature. But the vapour pressure 
which is in equilibrium with ice at a given temperature 
1 Discourse delivered at the Royal Institution on Friday, March 2, 1923. 
: ; humidity / 
is lower than that which’ is in equilibrium with super- 
cooled water at the same temperature ; that is, air is in 
equilibrium with ice at a relative humidity below roo 
per cent. 
Thus, according to our definition of relative humidity, 
the water vapour in air may be in equilibrium with 
water over a large range of relative humidities accord- 
ing to the physical state of the water present. The 
following tables give the relative humidity of air in 
equilibrium with pure water in the liquid and solid 
state. 
TaBLe I, 
Relative humidity of air in equilibrium with water 
surfaces of various radii at o° C. 



| | | : 
Radius, cm. | Flat. | roX ro. | 1X 10", | 2-5 X 10. | 7-2 X 1077, | 6X 107, 
| | | | 

Relative ) 

100 | TOO-OOO12 Ior 105 | 120 ) 400 

TABLE II. 
Relative humidity of airy in equilibrium with a flat 
surface of ice at various temperatures. 

| ] y 
Temperature | 0° C, | -- 10° C20" C.|—30° C.}—40° C.|—50" cael Cc. 




| 100 Sree 4 VS Bnd ee vie gs 
Relative \ 
humidity 

CONDENSATION AT TEMPERATURES ABOVE THE 
FREEZING Pornt. 
It was in 1880 that Aitken first showed that condensa- 
tion does not necessarily take place in air when its 
temperature is lowered below the dew point. He 
expanded carefully filtered air and found that no fog 
formed even when there was considerable supersatura- 
tion. Aitken was led to the conclusion “ that vapour 
molecules in the atmosphere do not combine with each 
other, that before condensation can take place there 
must be some solid or liquid nucleus on which the 
vapour molecules can combine, and that the dust in the 
atmosphere forms the nuclei on which the water- 
vapour molecules condense.” 
Aitken invented a most ingenious instrument, easy 
to work and very transportable, by means of which it 
is possible to count the number of nuclei present in the 
air. Thousands of tests of the atmospheric nuclei have 
been made with this instrument at many places over the 
world, and nowhere has air free from nuclei been found. 
The number of nuclei is seldom less than roo per c.c., 
while in most country places the nuclei rise to 
thousands, and in cities such as London and Paris the 
number may be so great as 100,000 to 150,000 per C.c. 
The general explanation of these observations is as 
P3 
