14 Dr. W. N. Shaw and Mr. W. H. Dines. [Apr. 28, 



only four stages of humidity are dealt with, and these are entered 

 upon the diagram, upon which are also recorded the observed heights 

 ofluds entered by the kites, the direction of the wind at the surface 

 and in the upper air, and particulars of the weather. 



For the purpose of comparison the curves of variation of the 

 barometer at Fort William and Ben Nevis during the period of 

 the experiments, are plotted on the same diagram and certain 

 particulars are also given about the temperatures of the wet and 

 bulb at those stations. 



From the diagram the fall of temperature for each 500 metres of 

 each ascent is taken out and tabulated. The table gives the following 

 average results. 



Table of Fall of Temperature in Degrees Centigrade for each 

 500 Metres of Ascent. 



July. August. 



Oto 500 metres 22 ascents 3'0 C. 13 ascents 2'6 C. 



500 1000 16 2-8 11 



1000 1500 9 9 



1500 2000 2 2-0 7 



2000 2500 1 ascent 2'0 3 



2500 3000 2 2-0 



3000 3500 2 1-7 



The range of fall for the first 500 metres varied from 4 C. to 1 C. The 

 smallest fall was associated with an inversion of temperature gradient 

 not far from the surface. An inversion of temperature gradient with 

 very dry air above a layer of clouds was shown also on one of the 

 occasions of steepest gradient near the surface. The steep gradients 

 observed in the lower strata are shown to be associated with anti- 

 cyclonic conditions preceding the approach of a depression, and by 

 examples on five occasions it is shown that the characteristic of the 

 passage of a depression is that the isothermal lines of the diagram open 

 out as the depression comes on, the average diminution of gradient 

 for the change of barometric condition amounting to as much as 

 50 per cent. 



The paths of the centres of depressions producing these changes are 

 shown on the maps taken from the monthly weather reports of the 

 Meteorological Office. It appears that they passed the station on all 

 sides at various distances but none actually crossed it. The results 

 show that whatever was the path taken by the centre, the column 

 of air over Crinan became relatively much more nearly uniform 

 in temperature under the influence of the depression, and therefore 

 probably represented a relatively warm column of air. 



The average of the values of temperature gradient in columns of 



