[ 69 ] 



accumulating the air incumbent on them in a greater degree 

 than the air incumbent over plains is condenfed at equal heights ; 

 and hence when the barometer on the plains falls, and that on 

 the mountain alfo, it will be found, after making allowance for 

 the difference of temperature, that the fall is proportionably 

 greater in the inferior than in the fuperior barometer; and on 

 the contrary, if the $ afccnds both in the inferior and fuperior 

 barometer, the afcent will be proportionably greater in the fupe- 

 rior than in the inferior. 



Thus General Roy on the 7th of Auguft, 1775, at nine 

 o'clock, found the correB height of a barometer on Carnarvon 

 quay 30,075, and on the peak of Snowden 26,418 inches. At 

 twelve o'clock that on the quay fell to 30,043, and that on 

 the peak to 26,405 ; the fall of the 2 on the plain was 

 therefore -j-^Vr of the whole, and the fall on the mountain was 

 only -rsVr of its original height. On the other hand, at two 

 o'clock, the barometer on the quay rofe to 30,045, while that 

 on the peak rofe to 26,415 inches corred height ; therefore that 

 on the quay afcended only -rrhrr of the whole, whereas that on 

 the peak afcended t^Vs- part of its height. Yet as the defcents 

 of the $ beneath its moft ufual mean height are much more 

 frequent and confiderable than its afcents above it, the varia- 

 tions on mountains are upon the whole proportionably fmaller 

 than at the furface of the fea. I am fenfible that fome obfer- 

 vations occur in which the 5^ has been found to fall on moun- 

 tains while it afcended on the plains ; but in all thofe I have 

 met, this happened in warm funny weather, on rocky fummits 

 which were heated in far greater proportion than their height 



demanded, 



