94 Lieut. 0. S. Clarke, On an optical method [Dec. 3, 



be calculated by means of the equation given. Two divisions 

 on the scale being thus obtained, the remaining divisions are 

 found by subdivision. 



Referring to Fig. 9 it will be evident that only at the in- 

 tersections and at one intermediate point between each pair will 

 circles traced round the drum be divided into parts, all of which 

 are equal. 



Circles traced at other positions will have one unequal division 

 at the line of junction, which will cause a slight abrupt move- 

 ment, or jump of the wave at each rotation of the drum. This 

 jump however cannot be mistaken for the steady progressive 

 motion which indicates that the position corresponding to the 

 stationarjr wave has not been arrived at. 



To make a reading it will be necessary to start the reed and 

 to move the hand-wheel h until a stationary double wave is seen 

 through the lenses e e, the pointer will then indicate the speed 

 of the drum. 



The reading can be made without taking the eye from the 

 lens, if a scale is placed close in front of the drum at such a 

 height that its graduations can be seen through the slit simul- 

 taneously with the waves. When such a scale is employed, it 

 is not necessary to bring the reed or fork close to the drum. 

 The prepared paper might in fact be wrapped round a distant 

 shaft, and the reading can be made through a small telescope in 

 the focus of the object-glass of which the slit is made to vibrate. 



In order to set tuning forks in vibration the arrangement 

 shown in Fig. 11 was devised, and has proved completely suc- 

 cessful. A piece of soft iron turned up at the ends is carried 

 on an axis between the prongs of the fork, the axis being turned 

 up to form a handle. By turning this handle till the soft iron 

 bar is in the position shown, the prongs are forced apart. If then 

 the handle is again sharply turned in the opposite direction the 

 fork is set in vibration. 



It has so far been assumed that the period of a fork or reed 

 is absolutely constant. This is not the case, as the fork varies 

 slightly with temperature, vibrating more slowly as the metal 

 becomes warmer. In some experiments made with tuning forks 

 a loss "Oil per cent, per 1" Centigrade v/as observed. This would 

 be too small to affect the value of the instrument for practical 

 purposes, while, if it were employed for delicate investigations, 

 a correction could readily be applied. Temperature similarly 

 affects reeds, their period is also lengthened by an increase of 

 pressure of the air by which they are • set in vibration. By the 

 employment of the air injector described in a foregoing paragraph, 

 any considerable variation in the pressure of the air supplied 

 is prevented. The mean of .22 fairly concordant observations 



