48 



MEMOIK3 OF THE NATIONAL ACADEMY OF SCIENCES. 



the mean of two successive secouds we have the mean number of vibrations for those seconds. 

 These means are given in column 3. 



Table I. 



(1) 



8 



9 



10 



11 



12 



(2) 



(3) 



255. 

 256. 

 255. 

 256. 

 254. 

 256. 

 2.54. 



254. 

 257. 

 254. 

 257. 



00 

 90 

 05 

 90 

 90 

 90 

 70 

 15 

 95 

 10 

 90 

 10 



255. 95 

 255.97 

 255. 90 

 255. 92 

 256. 02 

 256. 00 



(1) The mean of 1st and 2d secoDds= 255.95. Amplitude of vibration of 1st second ^2.03 millimeters. 

 The mean of Uth and 12th secouds = 256.00. Amplitude of vibration of 12th second = .63 uiillimeter. 



From this observation one might conclude that the number of vibrations increased with a 

 diminished amplitude, but the following observations show that this is not a just conclusion : 



(2) Mean Of 1st and 2d second8=255. 97. 

 Mean of 7th and 8th secouds=25.5. 97. 



(3) Mean of Ist and 2d seconds=256. 05. 

 Mean of 11th and 12th second8=256. 00. 



(4) Mean of 1st and 2d seconds=256. 17. 

 Mean of 9th and 10th seconds=256. 20. 



Amplitude of vibration of 1st second=l. 19 millimeters. 

 Amplitude of vibration of 8th secoud:= . 59 millimeter. 

 Amplitude of vibration of 1st second=2. 39 millimeters. 

 Amplitude of vibration of 12th 8econd= . 61 millimeter. 

 Amplitude of vibration of 1st 8econd^2. 07 millimeters. 

 Amplitude of vibratiou of 10th 8econd= . 78 millimeter. 



From the above measures we conclude that diflerences of amplitude of vibration in a fork, 

 arranged as in the experiments, has no appreciable effect on its vibratory period. 



Many measures were made on records obtained with varyiug pressures of the tracing style 

 against the smoked pajier; but the slight variations of those pressures which could be obtained 

 within the range of elasticity of the delicate style used gave no diflerences in the number of vibra- 

 tions from which we could detect any influence of varying pressures of the style. 



KPFEOT OF TEMPERATURE ON THE VIBRATORY PERIOD OF FORKS. 



To determine the effect of variations of temperature on the vibratory periods of steel forks, I 

 bought two sets of Kcenig's forks of the CT, harmonic series to known diflerences of temperature, 

 and then determined how much they were thus thrown out of unison by the ob.servation of the 

 number of beats thus can.sed in one minute of time. 



Instead of heiiting or cooling one set of the forks by automatic thermostats, which method 

 had several objections in principle and great diflicultits in the way of experimenting, I decided to 

 wait for a favorable spell of weather, which we often have in April, when the air is still and misty 

 and a drizzling rain occurs. In such weather the air is nearly constant in temperature. During 

 such favorable conditions for the work, when the atmosphere varied only a few degrees in tem- 

 perature during two days of mist and rain, I opened the windows of a room which contained one 

 of the sets of forks and allowed them to remain there for a night and part of a day before begin- 

 ning the experiments. In an adjoining room, kept at as nearly an equable temperature as 

 possible, I placed the other forks. After the respective temperatures of these rooms had not 

 varied perceptibly during three hours, I opened the door between the rooms just enough to hear 

 clearly tlie forks of one room when stationed near the forks in the other. The temperature of the 

 hot room was 66° Fahr., that of the other room was 41° Fahr. 



Simultaneously sounding in order the two corresponding forks of the series, I obtained the 



