ON TIMEKEEPEnS. 191 



face, a small increase of velocity will considerably increase the distance of 

 the weights from the axis, and consequently the effect of their friction, so 

 that the machine will be immediately retarded, and its motion may thus be 

 made extremely regular. It may be turned by a string coiled round the up- 

 per part, and this string may serve as a support to a barrel, sliding on a 

 square part of the axis, which will consequently descend as it revolves. Its 

 surface, being smooth, may be covered either with paper or with wax, and a 

 pencil or a point of metal may be pressed against it by a fine spring, so as to 

 describe always a spiral line on the barrel, except when the spring is forced a 

 little on one side by touching it slightly, either with the hand, or by means 

 of any body of which the motion is to be examined, whether it be a falling 

 weight, a vibrating chord or rod, or any other moving substance- In this 

 manner, supposing a barrel a foot in circumference to revolve in two se- 

 conds, each hundredth of an inch would correspond to the six hundredth 

 part of a second; and the scale might be still further enlarged if it were ne- 

 cessary. (Plate XV. Fig. 198.) 



By means of tliis instrument we may measure, without difficulty, the 

 frequency of the vibrations of sounding bodies, by connecting them with a 

 point, which will describe an undulated path on the roller. These vibrations 

 may also serve in a very simple manner for the measurement of the minutest 

 intervals of time; for if a body, of which the vibrations are of a certain 

 degree of fiequency, be caused to vibrate during the revolution of an axis, 

 and to mark its vibrations on a roller, the traces will serve as a correct index 

 of the time occupied by any part of a revolution, and the motion of any other 

 body may be very accurately compared with the number of alternations 

 marked, in the same time, by the vibrating body. For many purposes, the 

 machine, if heavy enough, might be turned by a handle only, care being- 

 taken to keep the balls in a proper position, and it would be convenient to 

 have the descent of the barrel regulated by the action of a screw, and 

 capable of being suspended at pleasure. 



But for the general purposes of timekeepers, all other inventions have 



•been almost universally superseded by the pendulum and the balance 



spring, or pendulum spring. About the year 1000, Ibn Junis, and the 



other Arabian astronomers were in the habit of measuring time, during their 



