2ae Proceedings of Indiana Academy of Science. 
long from a piece of steel shafting about 5 cm. in diameter. The 
shaft was arranged for rotation at a high speed inside a light-tight 
box Y connecting with another light-tight box X, with a rectangular 
opening O, 2x15 cm. between them. B: and B: are boxes to hold 
the full and empty spools for photographic film. Guides F on each 
edge of the film caused it to lie, when unwound, on the surface of 
a cylinder with the rotating mirror M on the axis. Just in front of the 
mirror is a horizontal rod R, of small diameter. A spark from the 
terminals E of an electric machine jumps the gaps Gi, G2, S and L, the 
spark at S occurring before the one at L. When the sound spark occurs 
the light passes through O, falls upon the mirror M, and is reflected on 
the film, the rod R producing a shadow R; on the film. Suppose that 
the sound pulse arrives at W, by the time the retarded light spark 
occurs. A part of the shadow of W; is intercepted by the wall of the 
box. A part passes through O, falls upon the mirror at W., W:2, and is 
reflected on the film at W:, W:, together with a second shadow of the 
rod R, now at R., due to the fact that the mirror has rotated through 
a measurable angle during the interval between the sound spark and the 
light spark. The distance between the shadows R: and R, together with 
the mirror speed and the distance from the mirror to the film, enable 
one to calculate the time interval between the sparks. 
From the measured distance W:.—W;, together with the distance from 
the light spark to the sound spark, and from the sound spark to the 
mirror, and thence to the film, one gets the true radius of the sound 
wave. The quotient of the radius by the time gives the average speed. 
If one plots radius by time for a considerable number of observations 
the tangent at any point on the curve gives the instantaneous speed at 
that point. 
The films used were eight inches wide and four feet long, and in- 
cluded about sixty degrees of the are about the mirror. As the image 
rotates twice as fast as the mirror it is evident that if the sparks were 
produced at random, there would be but one chance in twelve of the 
mirror being in the proper position to give a picture. To avoid this 
difficulty and to enable one to get several pictures on the same film a 
metal rod was fastened in such a position on the end of the mirror shaft 
that it shortened the gap G, to such an amount as to cause a spark to 
pass at the proper time. The position of the gap G. was varied by fixing 
the electrode J at different points along the arc A. J was arranged so 
it could be slid back and forth through a sleeve. When a spark was 
desired J was pushed forward and the gap G, thus shortened until a 
spark occurred. The gap was then lengthened before the electric ma- 
chine had time to generate a sufficient potential for a second spark. In 
practice, however, this device was found to be somewhat erratic, prob- 
