Short-Period Electrometer. 241 



For the purpose of measuring periods and amplitudes it is 

 convenient to have the lowest points of the curve in the zero 

 line. The condition that x should reduce to the form 

 Btf- 2 ^sin 2 .(m£ + <£) is 



(V - 2kfi + n 2 ) 2 = ( V - 4m 2 - 2kfi + n 2 ) 2 -f (fym - 2km) 2 , 



i.e. k 2 -2?i 2 -4kjjL + 4:m 2 + 8 l u 2 = 0. 



Unless m and \x are very small in comparison with k and 

 ??, the system must be adjusted to suit each case. The 

 adjustment may be made either by varying the mixture 

 of oil or, more conveniently, by varying the tension of 

 the strip until the lowest points of the curve are in the 

 zero line. If the tension is insufficient the spot of light does 

 not reach the zero line during the oscillations ; if the tension 

 is too great it goes beyond it. When the tension is accurately 

 adjusted, the ratio of the greatest ordinate in one wave to that 

 in the next but one should be <? 2 ^ T , where T is the period of 

 the applied electromotive force and ja its damping factor. 



(3) Method of starting the Oscillations. 



The electrical oscillations examined were those of a circuit 

 containing a coil, A, of large self-inductance (the secondary 

 of an induction-coil) and a leyd en-jar whose plates were 

 connected to the terminals of the electrometer. Two methods 

 were employed for producing the oscillations. In the first a 

 primary coil, with or without an iron core, was placed inside 

 A and a current in it broken as the spots of light were 

 crossing the photographic plate. For this purpose the 

 break, consisting of a vertical zinc rod attached to one end 

 of a lever and dipping into an iron vessel containing 

 petroleum, was placed underneath the rotating mirror. A 

 spring acting on one end of the lever usually kept the zinc 

 rod down in contact with the bottom of the iron vessel, but 

 when the mirror was rotating a plate attached to its axle 

 depressed the other end of the lever when the spots of light 

 were on the photographic plate and kept the primary circuit 

 open for about one-half of the period of rotation. PL IX. figs. 2 

 and 3 show two wave-curves obtained by this method. 

 The period of oscillation was in the first about *00488 second, 

 and in the second about '0113 second. The iron-cored primary 

 was used in each case, but in the first a condenser of about 

 03 microfarad capacity was connected across the break, and 

 in the second a condenser of about 1*8 microfarad was added 

 in parallel with this, making altogether 2*1 microfarad across 

 the break. 



