LOUDNESS 425 



The customary routine measurements to insure the proper voltage 

 levels impressed upon the receivers were made with the measuring 

 circuit shown schematically in Fig. 15. During the progress of the 

 tests voltage measurements were made frequently and later correlated 

 with measurements of the corresponding field sound pressures. 



Threshold measurements were made before and after the loudness 

 tests. They were taken on the same circuit used for the loudness tests 

 (Fig. 15) by turning off the 1000-cycle oscillator and slowly attenuating 

 the other tone below threshold and then raising the level until it again 

 became audible. The observers signalled when they could no longer 

 hear the tone and then again when it was just audible. The average 

 of these two conditions was taken as the threshold. 



An analysis of the harmonics generated by the receivers and other 

 apparatus was made to be sure of the purity of the tones reaching the 

 ear. The receivers were of the electrodynamic type and were found 

 to produce overtones of the order of 50 decibels below the fundamental. 

 At the very high levels, distortion from the filters was greater than 

 from the receivers, but in all cases the loudness level of any overtone 

 was 20 decibels or more below that of the fundamental. Experience 

 with complex tones has shown that under these conditions the con- 

 tribution of the overtones to the total loudness is insignificant. 



The method of measuring loudness level which is described here has 

 been used on a large variety of sounds and found to give satisfactory 

 results. 



Appendix B. Comparison of Data on the Loudness 

 Levels of Pure Tones 



A comparison of the present loudness data with that reported 

 previously by B. A. Kingsbury * would be desirable and in the event 

 of agreement, would lend support to the general application of the 

 results as representative of the average ear. It will be remembered 

 that the observers listened to the tones with both ears in the tests 

 reported here, while a single receiver was used by Kingsbury. 



Also, it is important to remember that the level of the tones used in 

 the experiments was expressed as the number of db above the average 

 threshold current obtained with a single receiver. For both of these 

 reasons a direct comparison of the results cannot be made. Howev^er, 

 in the course of our work two sets of experiments were made which 

 give results that make it possible to reduce Kingsbury's data so that 

 it may be compared directly with that reported in this paper. 



In the first set of experiments it was found that if a typical ob- 

 server listened with both ears and estimated that two tones, the 



