The Objective Reality of Combination Tones. 397 



The resonator they employed was a tuning fork. On to one prong 

 of the fork was fixed a mirror, which was made one of a system by 

 which Michelson's interference bands were produced. To the other 

 prong was attached a wooden square of larger area, but of the same 

 weight as the mirror. The fork was then compared with a Kouig 

 standard fork and its frequency adjusted to 64. The notes were 

 produced by a Helmholtz wind siren placed between a wooden 

 pyramidal tube and a large Konig resonator tuned to 64. The 

 narrow end of the pyramid was placed about half an inch from the 

 wooden square attached to the resonating fork. 



Throughout the experiments we used blue light, obtained by 

 passing a beam from an electric lamp through a cell containing an 

 ammoniacal solution of copper sulphate. A slit about 2 inches 

 long and one-twentieth of an inch broad was cut out of a piece of tin- 

 foil pasted on glass, and was placed horizontally across the middle of 

 the bands, so that the bright and dark bands appeared as bright and 

 dark spots respectively. 



In our earlier experiments, we took photographs upon flexible film 

 fastened by india-rubber bands to a rotating drum. A ball shutter, 

 such as is used in instantaneous photography, was employed. Working 

 at night, after the traffic had subsided, the bands in their normal 

 condition were perfectly steady. On taking a photograph, each 

 bright band produced a perfectly straight line upon the rotating film, 

 and the whole picture was made up of a series of parallel straight 

 lines. 



On sounding a 64-fork in the vicinity of the apparatus, the 

 mirror is set in motion by resonance and thus the bands execute 

 harmonic motions, with a frequency of 64, about their mean positions. 

 The general appearance as then seen by the eye is a blue blur. When 

 this blur was photographed by means of the slit and rotating drum, 

 we obtained a series of sinuous lines. To prove that the frequency 

 of these curves corresponded to a note of 64 vibrations per second, we 

 made observations on the rate of turning of the drum. The following 

 are some of the actual figures obtained : 



Circum. of drum = 31*5 cm. 100 revs, in 19'5 sees. 

 Wave-length = 2'53 cm. 



No. of vibns. per sec. = ^9.5 x 2-53 = ^3'9. 



We thus obtained a series of photographs of what may be called 

 the difference arid summation curves, which were exhibited in a 

 preliminary communication read by Professor Riicker at the recent 

 meeting of the British Association held in Toronto. 



The details of the method now employed are different, and we 



