Study of Locomotive Whistles 109 



of blanket over the locomotive which still further absorbs and disperses 

 the sound waves produced by the whistle below. 



Another factor which had to do with the sound distribution shown 

 in figure 1 is the design of the whistle itself. The usual cylindrical tube 

 forming the resonator (bell) of a single tone whistle is, in the case of 

 the chime whistle used in this study, divided by longitudinal radial vanes 

 into five compartments or pipes, each of the proper length to give one 

 of the notes of the chime, viz, C, E, G, C and C". T in figure 4 is a 

 transverse section of the whistle and L a longitudinal section, with the 

 omission of the valve mechanism at V. The former shows the relative 

 positions and cross sectional areas of the five pipes while the longitudinal 

 section shows the relative lengths of two of the pipes — C and its octave 

 C. The fraction of the cylindrical steam jet J used in blowing each 

 of the pipes is shown in the transverse section T, and was 26 per cent 

 in the case of the lower tone C, and respectively 22, 19, 17, and 16 

 per cent for the other four tones. Thus 60 per cent more energy was 

 used in blowing the lower tone than in blowing the upper tone of this 

 whistle. Since the quality or character of a sound depends on the 

 relative intensity of the several tones which combine to form it, it is 

 evident that the quality of the sound from a chime whi.stle depends to 

 a degree upon which pipe of the whistle is toward the observer. This 

 variation with direction is accentuated when the whistle is placed very 

 near a steam dome, which interferes more or less with the normal func- 

 tioning of that part of the whistle which happens to be nearest it. 

 Inasmuch as the Rayleigh disk was not equally sensitive to all five tones 

 of the chime, intensity measurements made with the disk showed con- 

 siderable variation whenever there was a change in the orientation of 

 the whistle with respect to the dome. 



The writer would locate a locomotive whistle in front of the loco- 

 motive where it would be free from the several disturbing- factors 

 named. He would place it in a reflector of such design as to give a 

 maximum of sound intensity ahead of the locomotive and of such size 

 as to serve as a resonator and thus increase the intensity of the sound 

 at the source. 



In advocating the use of a reflector to direct the sound of the 

 whistle along the track the writer has continually met with the argu- 

 ment that such a device would be practically useless on account of the 

 fact that the reflector could not be made large compared to the length 

 of the sound waves to be reflected. This limitation does, of course, 

 affect profoundly the rate at which sound wave energy spreads out 

 after the waves are outside the reflector. But it does not change the 

 action of the reflector itself. The energy of the reflected portion of 

 the sound wave can be headed in the right direction, and much of it will 

 continue in the right direction to reinforce the wave originally pro- 

 jected in that direction. In proof of this assertion it is sufficient to 

 call attention to the eff'ectiveness of a megaphone as a sound director. 

 We should expect a relatively greater directive action from a reflector 

 placed about a whistle. In the case of the voice the waves are originally 

 projected in one hemisphere only, the whistle starts them in both hemi- 

 spheres. The whistle reflector reflects a portion of the waves in the 



