THE USE OF A BALLISTIC GALVANOMETER AND A PENDU- 

 LIBI FOR :\IEASURING RAPIDLY FLUCTUATING 



RESISTANCES. 



BY W. H. CLARK. 



The degree of accuracy with which physicists have divided the funda- 

 mental units of length and mass, the centimeter and the gram, demands 

 a division equally as accurate of the remaining unit, the unit of time. 

 We are all familiar with ]\Iichelson 's measurement of the standard meter 

 where a half and a quarter wave length of light was a quantity not to be 

 neglected. Our recent literature describes a balance which will weigh 

 to one millionth of a milligram. When we consider these facts and the 

 added fact that time so often enters into our formulae to the second de- 

 gree we see the necessity of accurate small divisions of our unit of time. 

 For special purposes such divisions have been made illustrated by the 

 many forms of chronographs. The kind of time measuring instrument 

 is usually determined by the purpose required. There is the gun chrono- 

 graph, the astronomical, the accoustic and the chronograph for physi- 

 ological purposes. All of these are more or less complicated in their 

 structure. An instrument exceedingly simple in design and still suffi- 

 ciently accurate for the purpose has been used in the physical laboratory 

 at the State University of Iowa to measure a rapidly fluctuating resist- 

 ance. It is a ballistic pendulum which swings over an arc which is 

 graduated with respect to time. The shortest time which could be meas- 

 ured was .00125 second. Keys which opened and closed electrical cir- 

 cuits were placed along the arc at any desired points. These were tripped 

 by the pendulum. A variable resistance was placed in one arm of a 

 Wheatstone bridge circuit. We found that its average value for iny 

 interval of time can be expressed in terms of the throw of a ballistic gal- 

 vanometer where the galvanometer is thrown into the circuit by means 

 of the pendulum for the desired interval of time which must be small 

 compared with the period of the galvanometer. How small this in- 

 terval should be is also determined by the rapidity with whicli the re- 

 sistance changes. If the time is sufficiently short the average resistance 

 practically becomes the actual resistance at the middle of the interval. 



