792 MR J. Y. BUCHANAN ON THE 



maximum and minimum have been nearly smoothed out, and the curve changes char- 

 acter with the approach of winter. The most striking feature of these curves is the 

 sharp rise of pressure between midnight and 1 a.m., this is apparent in all the curves 

 from November to April, with the single exception of December. The morning minimum 

 is found about 6 a.m., with a forenoon maximum at 11 a.m., and these are followed by 

 an afternoon minimum and an evening maximum more or less pronounced. The extra- 

 ordinary serration of the curve after midday in April is remarkable, as it is the only one 

 belonging to foggy weather which shows this feature. * 



A general summary of the movements of the barometer during the year is presented 

 in the tables on the precediug page, in terms of the inch and the millimetre 

 respectively. 



Tension of Aqueous Vapour. — This element is dealt with in Tables XX. to XXVI. 

 Tables XX. and XXI. contain the absolute maximum and minimum values observed 

 at every hour in the thirteen years in clear and in foggy weather respectively. Tables 

 XXII. and XXIII. contain the monthly mean hourly values of the aqueous tension ; 

 Tables XXIV. and XXV. contain the diurnal variation of the same element ; and 

 Table XXVI. contains the monthly mean hourly excess of vapour tension in clear 

 weather over foggy weather. Excepting in the middle of summer this difference is 

 always negative, the aqueous tension in the colder part of the year being always higher 

 in foggy than in clear weather. 



In foggy weather the vapour tension is that of saturation at the temperature of the 

 air. As the water particles which form the fog permeate the whole of the air any addi- 

 tion of heat has the effect of changing into vapour a portion of the fog, and any loss 

 of heat has the opposite effect, of condensing some of the vapour of the air on the water 

 (or ice) of the fog, both being accompanied by a certain change in temperature. In the 

 clear weather the atmosphere is generally in a state far removed from saturation, and 

 the effect of addition or removal of heat is simply a rise or fall of temperature. In itself 

 the air in clear weather has no means of increasing its supply of water vapour. 



It follows from this that in the case of non-saturated air the percentage of aqueous 

 vapour present in it forms a valuable means of identification. If it has no means of 

 receiving or parting with water this percentage must remain the same. If the 

 percentage has changed, then we conclude that the whole air has changed either by 

 mixture or by replacement. Hence the percentage of aqueous vapour affords a ready 

 means of detecting changes of air. 



The barometric pressure is the measure of the sum of the tensions of the various 

 gaseous constituents of the atmosphere. If this pressure is diminished all the constituents 

 expand proportionately, and their tensions diminish also proportionately. If the pressure is 

 increased they contract in the same proportion, and their tensions increase proportionately. 



The tension of a given quantity of gas varies inversely with the volume which it 

 occupies. From this it follows that, if, in a gaseous mixture, we know the total tension 

 of the mixture, which is given by the barometric pressure, and the tension of any one 



