MEASUREMENT OF NOISE KAYE 



187 



istics were among the probable contributory factors. It was not 

 found possible to correlate stridency with wave form. The spectra 

 and wave forms were obtained of a particular electric klaxon with 

 alternative types of noise, an electric buzzer, an English bulb horn 

 (reed type), and a French bulb horn (reedless). In some of these 

 cases, the characteristics were substantially modified by adjustment 

 of the horn. 



The New York Noise Commission arrived at much the same con- 

 clusions. They regard horns with sound levels in excess of about 90 

 decibels when heard 23 feet away, as unnecessary and objectionable. 

 They find that complaints of stridency are unlikely to arise when 



80 



in 





^ 



60 



20 



r28 2SS 582 ID24 2048 4096 8192 



Fref|,ucncu (cibcIcs per sec) 



Crowd Moises measured IJO<|eel* cswau. 



FlUUKE !.l 



fundamental frequencies lie between 200 and 300 cycles, when the 

 overtones are all harmonics of the fundamental and share the energy 

 evenly, and when there is an absence of strong high frequencies. 



As regards trains, the noise levels of express and suburban trains 

 in England and America seem to be not unlike for a similar class 

 of accommodation. The American method of dividing up Pullman 

 sleeping-cars partly by means of heav}'^ curtains seems, despite its 

 other drawbacks, to result in a noise level comparable with that of 

 our own more secluded first-class sleeping berths. In such circum- 

 stances, however, much depends on other factors, such as the good 

 fitting of doors and windows, which restrict noise admission. 



With reference to underground railways, the New York subway 

 stands in a class apart for noise — as anyone who has traveled by it 

 will testify. Our own tubes appear to be at least 10 decibels quieter, 

 though questions of speed may come in. 



