86 Professor Potter on the Aerometric Balance, 



Here m and n are small quantities, but m much smaller 

 than n, so that 





y 



, r 



= 1 + 



X 









_ W 2 





-n) 

 —m) 







w n 



.r 1 





w, .r 



(l — n — m) nearly 



^.rl \c b) J '"' 



and the value of x will be continually changing as the value 

 of «, the specific gravity of the air, changes, whilst the frac- 



tion 2 * remains constant, and the changes in b and c are 



w 1 .r to 



almost imperceptible in comparison with those of a : or 0, with 

 the position of equilibrium, will continually change with the 

 density of the atmosphere. 



The degree of sensibility may be any that is wished, pro- 

 vided weights are added to, and taken away from the end B, 

 which weights might be very conveniently laid upon the plate 

 containing the agate plane. In some trials I had the index 

 moving through 40° with i grain difference in the weight 

 hanging from B ; and this is equivalent to a change of about 

 one inch in the height of the barometer, or between 16° and 

 17° of Fahrenheit's thermometer. 



When changes in the density of the air are required to be 

 found which are not dependent solely on temperature and 

 pressure, but also on occasional admixtures in the air sur- 

 rounding the balance, we have to compare its indications with 

 those of the other named instruments, and it is needless to 

 have a greater sensibility than theirs. Now the best baro- 

 meters read only to one-thousandth of an inch in the height 

 of the mercurial column, and the best thermometers can 

 seldom be depended upon to one-hundredth of a degree 

 Fahrenheit; so that if ^ a grain added at B cause the index 

 to move through 10° of arc, and we read the vernier to \ mi- 

 nutes, we have an accuracy greater than either of the other 

 named instruments, and one which is sufficient for our expe- 

 riments. In this case the scale of 60° will suffice for the 

 changes which occur in this climate, without need of weights 

 in addition. 



