MEASURING THE VIBRATORY PERIODS OP TITNINd-FORKS. 



55 



422. 793 



Table XY. 



[Marloye (C) fork.] 



Tbe Ueterniiuatious of the number of vibratious per second of tlie .4 forks have to be cor- 

 rected by +.04-4 for the cfl'ect of the weight of the tracing style. The correctiou was too small 

 to be determiiied iu the case of the C fork. 



The separate detei'miiiations of the number of vibratious of the Oiiapelle Versailles fork and 

 those of the Theatre Feydeaii fork are numerous enough to give some idea of the probable error 

 of a single determination and of the error of tbe mean of the determinations when these are dis- 

 cussed by the method of least squares. 



From this discussion it appears that for the Chapelle Versailles fork, the probable error of a 

 single determinations: i .019 of a vibration ; the probable error of the mean determination= + .0053 

 of a vibration. 



From the experiments on the Theatre Feydeau fork, the probable error of a single determina- 

 tion= i.014 of a vibration; tbe probable error of the mean determination^ ±.004 of a vibration. 



These results show tliat the method is quite accurate, and certainly sufidciently so for the 

 determination of the pitch of a standard fork, and for all purposes when the fork is used as a 

 chronoscope iu the measure of small intervals of time. If the error of the determination of the 

 pitch of these two forks — when corrected for eflects of support and S(;rape, which is a constant 



