520 Mr. H. S. Allen on the Motion of 



employing photography in the determination of the velocity 

 might be suggested, but the plan finally adopted was as 

 follows. 



A photographic plate was placed at the back of the rect- 

 angular glass vessel in which the balls were allowed to fall. 

 The plate was illuminated by a series of flashes of strongly 

 actinic light emerging from a small aperture at some distance 

 from the front of the glass vessel. In this way the shadow 

 of the ball was thrown upon the plate by each successive 

 Hash, giving a permanent record of its position at the corre- 

 sponding instant. In order to secure a succession of flashes 

 at equal intervals of time, twelve equidistant radial slits were 

 cut round the circumference of a disk, 20 centim. in diameter, 

 kept revolving at a uniform rate ; in the course of a revolution 

 each slit was brought opposite a fixed vertical slit on which 

 a beam of light was concentrated by a short-focus condensing 

 lens. It is clear that this arrangement would give a series of 

 images of the ball in a vertical line on the plate, but any one 

 image would be fogged by the light from the remaining 

 flashes. To prevent such a result a metal screen having a 

 vertical rectangular opening was placed between the plate and 

 the back of the glass vessel, and the plate was drawn hori- 

 zontally past this opening so as to expose a fresh surface to 

 each flash. The glass vessel, screen, and plate were inclosed 

 in a wooden case, in the front of which was a circular aperture 

 closed by a Thornton-Pickard photographic shutter to limit 

 the duration of the exposure. 



It will be seen that the degree of success attainable in the 

 measurement of the velocity depends in the first place on the 

 constancy of the rate of rotation of the revolving disk and 

 our ability to measure that rate, and in the second place on 

 the sharpness of the images on the developed plate. 



The first requirement was satisfied with a very high degree 

 of accuracy by attaching the disk, with its twelve radial slits, 

 to the fly-wheel of the modification of Froment's electromag- 

 netic engine devised by Lord Rayleigh*. The motor was 

 driven by a current from storage-cells rendered intermittent by 

 a tuning-fork interrupter making about 30 complete vibra- 

 tions per second. The speed of rotation was obtained by 

 means of a counting-wheel geared to an endless screw on the 

 axle of the motor. 



In 36 min. 7 sees, the counting- wheel made 197 revolu- 

 tions. This gives exactly 11 sees, as the time of one 

 revolution. The counting-wheel possessed 45 teeth, and since 



* Phil. Trans, clxxiv. pp. 316-321 (1883). 



