1895.] on the Absolute Measurement of Electrical Besistance. 611 



arrangements we can obtain any required rate of rotation from about 

 150 to 1500 revolutions per minute. 



It is of the greatest importance, if the observations presently to 

 be described are to be made with ease, that the rate of rotation should 

 remain constant for the four or five minutes required, and no effort 

 made to compass such uniformity is thrown away. The design, work- 

 manship, and lubrication of all the bearings require closest attention, 

 and the friction between the brushes and the motor commutator should 

 be reduced to a minimum. In my Cardiff apparatus the axle near 

 the motor bears a heavy fly-wheel. 



During a run the observer controlling the rate of rotation requires 

 some test of uniformity, so that he may, if there is a quickening or 

 slowing down, diminish or increase respectively the current through 

 the electromotor so as to recover the initial sj^eed. 



To accomplish this it is convenient to refer the rate of rotation 

 by a stroboscopic method to a suitable tuning fork provided with 

 riders and maintained in vibration electrically. Here is such a fork. 

 At the end of each prong there is an aluminium plate with a slit in 

 it parallel to the prong. When the fork is in equilibrium the slits 

 are opjDosite one another, and an observer can see through both. 

 When the fork is in vibration he can see through the slits only when 

 they are opposite one another. This position occurs twice in every 

 complete vibration. Hence, if the number of vibrations per second 

 is, say, 64, he obtains 128 views per second. 



Now let us suppose that he looks through the vibrating slits at 

 a drum or disc coloured black, on which a number of white spaces, 

 which I will call teeth, are painted at equal angular intervals. Here 

 is such a disc attached to the rotating axle ; on it there are three 

 circular rows of radial teeth. 



When the disc is in rotation, if the rate of rotation is such that 

 in the interval between two views a tooth in any row exactly takes 

 the place of the next, that row will appear, to an observer looking 

 through the fork slits, stationary ; if it does not quite do so, the row 

 will appear to be moving backward ; if it passes beyond, the row will 

 appear to be moving forward. 



These phenomena I can show you by illuminating the rotating 

 disc by intermittent light. The light of this alternating arc varies 

 in intensity at each alternation, passing from maximum to maximum 

 through a minimum that is, however, very far from darkness. Our 

 speed is such as to make the row of fourteen teeth appear almost 

 stationary in this intermittent light. You can see the teeth, but they 

 stand out somewhat faintly. The faintness is due to the fact that we 

 have not complete darkness between the maxima of illumination. 



We may do much better with such a tuning fork as I have 

 described. I place its slits in front of this electric lantern, in which 

 there is a continuous arc. The light falling on the disc has passed 

 through the slits of the vibrating fork, and by varying the rate of 

 revolution we obtain perfectly distinct stationary teeth. Let us 



