[ 103 ] 



Example of the above Procefs. 



On conftrudlng a barometer, with a cylinder of one inch in 

 diameter, I obferve the mercury to ftand at 29, o, I open the 

 cyHnder, and pour in three inches from a piece of the refervcd tube, 

 replacing the whole as before ; I obferVe that the mercury now 

 ftands at 29, I2. It is obvious, from this experiment, that had 

 I carried the barometer, previous to this infufion of mercury, 

 to an height which would have occafioncd the defcent of three 

 inches of mercury into the ciftern, that the effed would have 

 been precifely fimilar, i. e. that the mercury would then fland 

 at a6, 12; and of courfe, in eftimating all heights, the interval 

 found by this barometer is in the firfl inftance to be Increafed 

 in the proportion of ,04 of an inch for every inch the mercury 

 falls in the tube. It is eafy to apprehend that a table may be 

 readily formed to make thefe allowances at fight, as corred as 

 can be defired.- 



On the advantage of this adjuftment it is to be obferved, that 

 being once found, it remains perfedly free from any pofTibility 

 of alteration, as neither the form or capacity of the tube or 

 ciftern are liable to alter. It faves the trouble of pouring 

 mercury in or out of the ciftern ; the inconvenience of leather 

 bags, that are conftantly liable to go out of order and wafte the 

 mercury ; and, befides, gets rid of all the errors of adjuftment, 

 arifmg either from fridion or reading ofFj which the obferver Is 



liable 



