[ 45 ] 



Every part of the inftrument is now fuppofed intirely fil- 

 led for carriage ; but when an obfervation is to be made, fome 

 of the mercury muft be let out, that the remainder may def- 

 cend to the ftandard altitude ; and for this purpofe there is 

 a hole at D, PI. ifl:, fhut by the ftop-cock E ; turn this cock 

 with the crank, and the mercury will run out until it comes 

 to the original level ; then, having turned the cock E back 

 again, take out the pieces C and F, and regulate the gage by 

 the plug A, as before, and the obfervation is compleated. 



This barometer, like all others, fliould be truly vertical at 

 the time of obfervation, and as this inftrument ftands upright 

 on a table refting on its bafe, if to this three adjufling fcrews 

 are attached, and a circular level on its upper furface, or crofs 

 levels, the horizontal ftate of the box is eafily afcertained; 

 but if any fufpicion arife that the tube may not be at right- 

 angles to the top of the box, then the inftrument may be hung 

 in the ufual manner in gimmals. 



The ftrongeft objeftions that occur to me againft the ufe of 

 this barometer, are the trouble of carrying a fraall phial of 

 mercury, and the poffibility of air mixing with the mercury 

 each time it is poured out at the hole D, but the trouble is 

 not worth mentioning ; and as to the air, it may and ought to 

 be expelled at every obfervation, by heating the mercury in a 

 ladle or crucible, and the accuracy refulting in very nice ex- 

 periments will repay the trouble. 



It 



