to the SURFACES of BODIES. 33 



hand, prevloufly ruled for that purpofe. And thus we could 

 with eafe mark the precife time of every revolution. The num- 

 ber of revolutions was ufually continued to about 35 ; and as 

 the motion, with the hemifphere commonly became uniform 

 after two or three revolutions, and with the lead only, after 

 about 20 revolutions, I fubtraded the time of the firft 25 revo- 

 lutions from that of 35, and the remainder was the mean time 

 of 10 revolutions ; and, confequently, dividing by 10 gave me 

 the mean time of one revolution very corredlly ; and thence, 

 from the fpace of one circle or revolution, which is 27.93 feet, 

 the velocity of the hemifphere per fecond of time. 



9. In this manner, then, by varying the actuating weight, 

 by I dr. or 2 dr. i^c. at a time, I obtained a long feries of cor- 

 refponding times and velocities, both with the round and flat 

 fide of the hemifphere foremoft, and with the equivalent lead 

 only. After which I fubtraded the numbers of this latter from 

 the correfponding ones of the two former, and the remainders, 

 when divided by 51.14, gave the true meafure of the preflure of 

 the air at the centre of the hemifphere. 



In the following table are feleded only the velocities in 

 whole numbers of feet, namely, of 3 feet per fecond, of 4 feet, 

 of 5 feet, and fo on to that of 20 feet per fecond of time, with 

 the correfponding adluating weights in all the three cafes, name- 

 ly the flat fide foremofl:, the round fide foremoft, and without 

 the hemifphere, with the lead only ; namely, fuch adluating 

 weights as were really experimented, and before dividing them 

 by J 1. 1 4, to reduce them to the centre of the body. 



Vol. II. c A B S- 



