346 Mr. Scrymgeour's Experiments to determine the Influence 



in the times of the vibrations, whether long or short. The 

 explanation of this fact is, that the retardation arising from 

 the current as the vibrations decreased, was equivalent to the 

 acceleration resulting from the short vibrations; that is, that 

 the current was stronger in proportion to the smaller vibra- 

 tions, and likewise to the diminution of the momentum of the 

 moving mass. 



The difference in time between the long vibrations in air 

 and in vacuo is less than it is between the short vibrations in 

 air and in vacuo. The reason of this is, that as the velocity 

 increases, the current is not so effectually formed in the long 

 as in the short vibrations ; besides, the air in front of the pen-* 

 dulum may be supposed partially condensed; and that behind 

 it partially rarefied ; — causes which both operate in producing 

 a quicker return of the pendulum. Indeed I am of opinion, 

 that if the vibrations were extended to 4° or 5°, they would 

 be faster in air than in vacuo, at the same extent of vibration. 



In the next experiment, the knife-edges were removed and 

 replaced by portions of a cylinder, of which the diameter was 

 1*8 inch, in such a manner that the distance between them 

 was the same as that between the knife-edges. The pendu- 

 lum, which weighed 9 pounds 10 ounces, was adjusted so as 

 to vibrate in the same time, when either the light or the heavy 

 end was lowest. In this state, the clock gained upon the pen- 

 dulum one vibration in 15 minutes, at the mean extent of 

 1°*3 vibration. The weight of the pendulum was then re- 

 duced to 4 pounds 8 ounces, and adjusted as before; the clock 

 now gained one vibration in 13 minutes (being a mean of se- 

 veral experiments,) at an extent of 1°*3 vibration, either with 

 the light or the heavy end lowest. These results give a dif- 

 ference, in the first experiment, of 1 minute 36 seconds in 24 

 hours, and in the second, of 1 minute 50 seconds, between the 

 times of the vibrations on the knife-edge and the cylinder, those 

 on the latter being slowest. The vibrations of the pendulum 

 in the light state were also slower than in the heavy state ; 

 this result was occasioned by two circumstances ; the current 

 had more effect on the pendulum, because in the former state 

 it presented a greater surface to the opposing medium, in pro- 

 portion to its weight, than in the latter state ; and because, in 

 order to produce the adjustment, it was necessary that the 

 weights should be nearer the end of the rod, when vibrating 

 on the cylinders, than when vibrating on the knife-edges. 



By experiments with cylinders of different diameters, I 

 found that in order to make a simple pendulum, which vibra- 

 ted seconds on a knife-edge, vibrate seconds likewise on a cy- 



