88 THE ROYAL SOCIETY OF CANADA 
It should be noted that the custom of comparing hygrometers, 
or humidities by any method based on the assumption that two samples 
or regions of air examined at about the same time within a few feet 
of one another, will have the same humidity, is often most unreliable. 
The distribution of water vapour in a well-ventilated laboratory is 
usually far from homogeneous. The presence of free surfaces of 
liquid, the arrival of the observer, and the convection currents in the 
air may produce a quite appreciable variation. Comparisons should 
be made with prepared volumes of air of known moisture content. 
The chief disadvantage of this method lies in the possibility of 
mixing the contents wrongly by a careless manipulation of the stop- 
cocks and mercury. Labelled valves, however, should safeguard 
the hygrometer from this accident at the hands of a novice. 
§ 3. A SIMPLE WEIGHING HYGROMETER. 
If a solution has an aqueous vapour pressure which is less than that 
of the air in contact with it, there will be an absorption of vapour 
until the two pressures become equal; and, if on the other hand its 
vapour pressure is greater, it will evaporate until the pressures are 
equalized. Solutions of highly deliquescent salts with a large amount 
of well ventilated surface per unit volume of solution were found 
to vary in weight with the atmospheric aqueous vapour pressure 
in a regular and sensitive manner, and a simple hygrometer based 
on this principle was constructed accordingly. 
Unfortunately no general and accurate formula appears to have 
been obtained for connecting the aqueous vapour pressure with the 
concentration of a solution. Such expressions as log, . = x ate 
almost useless if, as in the present case, there is any appreciable 
variation in the total volume of the liquid. A series of comparisons 
using prepared samples of air can, however, be made with an absolute 
hygrometer, and a chart or scale prepared which gives the vapour 
pressure at once for any indicated weight. 
Fig. 2 is a photograph of a hygrometer prepared in this way. 
A solution of phosphorus pentoxide was used, it having first been al- 
lowed to absorb water vapour until its vapour pressure was equivalent 
to about 1 cm. of mercury. (Solutions of calcium chloride, cobalt 

*f =vapour pressure of solvent at given temperature. 
{’ =vapour pressure of solution at given temperature. 
n =number of gram-molecules of solute. 
N =number of gram-molecules of solvent (as vapour). 
