202 Proceedings of Indiana Academy of Scieyice 



tions were made at night, for noise during the day interfered with the 

 hearing of the extremely low tones. All the curves shown have been 

 checked on at least three different evenings and I am certain that the 

 slope and relative intensities at the different distances are as nearly 

 correct as the conditions of the experiment and the apparatus would 

 permit. 



Data: 



Above is the data for the four curves shown in figure 4. D is the 

 distance from the mouth of the horn to the nearest end of the singing 

 tube. R is the resistance "plugged in" in the test circuit A to reduce 

 the tone to the same intensity as that in the reference circuit A'. 



It is noticed there is a marked superiority in the Foley Telephone 

 Mouthpiece (parabolic horn) as is shown by the curves 3 and 4 (fig. 4), 

 and that this superiority holds especially at the ordinary speaking dis- 

 tances of from 5 cm. to 10 cm. It was necessary to plug in almost 

 twice as much resistance in the circuit at 5 cm. to reduce the intensity 

 of the parabolic horn to minimum audibility as for the No horn, and 

 the ratio is about 5 to 1 for the flared horn. This increased efficiency is 

 probably due in no small part to resonance" in the horn and is as 

 marked for the speaking voice as for the pure tone of the singing tube. 

 When the small disc was inserted in the mouth of the horn so that the 

 focusing effect was a maximum the efficiency was considerably increased 

 for 4 cm. to 9 cm. as is shown by cuive 4. This curve exhibits a freak- 

 ish tendency to rise upward until a distance of about one centimeter is 

 reached when the slope reverses and it drops off in a smooth curve. This 

 tendency of the curve to rise for a short distance was probably due. 

 to the shielding effect of the disc when the source is near to it. The 



•^ D. C. Miller. The Science of Musical Sounds, pp. 1.56-159. 



