APPLICABLE TO MATHEMATICS, &C. 335 



moves from one point of the barometer to 

 another should be noticed, and no doubt there is 

 a relation between the time in which the variation 

 of the barometer takes place, and the time of 

 transmission of fluid pressure through the atmos- 

 phere, from one point to another. The observa- 

 tion of the element of time, readily suggests 

 itself from the circumstance that no alteration 

 can take place in the weight, and consequently 

 in the pressure of the atmosphere, only by means 

 of adding to, or taking away material particles 

 from it ; this appears to be so much in accordance 

 with our notions of the doctrine of abstract forces, 

 that few deny the truth of it. It is true, that if 

 the variations of the weight of the atmosphere is 

 the cause of the fluctations of the barometer, we 

 ought to have the mercury falling during the 

 time that the rain is descending ; this is not 

 always found to be the case, the barometer some- 

 times rises during the time that rain is falling ; 

 this circumstance may be produced by the dis- 

 disturbance of the atmosphere at a considerable 

 distance from the place where the barometer is 

 situated. 



Additions to the weight of the atmosphere at 

 any place on the earth's surface can only be made 



