OF ARTS AND SCIENCES. 67 



It was also noticed that the vowels were broadened, a becoming a, etc., 

 the effect of the tight membrane in both cases being the strongest. 



Musical sounds passed through a tiglit membrane appeared mudi 

 louder, and those of the highest pitch much more intense, than when 

 passed through the open mouth-piece. While experimenting with 

 sounds of a low pitch passed through a tight membrane, it was noticed 

 that, while the fundamental tone was hardly perceptible, the first har- 

 monics could be phiinly heard. This fact seems to have the following 

 explanation : the membrane being found to be much more sensitive to 

 sounds of a higher pitch than to those of a lower, when the low tone 

 was sounded the membrane transmitted the vibrations of the har- 

 monics sounding with it, of a higher pitch than the fundamental tone, 

 before the fundamental tone itself. Without the membrane no such 

 effect was produced. Loose membranes gave the same effect, though 

 relatively to the experiments with the voice much less intense, being 

 weaker than the sound transmitted through the open mouth-piece. 



While trying the above experiments the idea suggested itself that 

 two telephones placed end to end, having one diaphragm in common 

 and furnished with two mouth-pieces connected by long tubes, would 

 furnish an instrument for measuring reflected sound. Such an instru- 

 ment could be used as follows. Place one of the mouth-pieces to receive 

 the direct and the other the reflected sound. Now, if the two sounds 

 were equal, nothing would be heard at the telephone placed in the cir- 

 cuit for the observer, as proved by the result of the experiment made 

 in transmitting one sound through two mouth-pieces. If one sound 

 was stronger than the other, its vibrations would be transmitted, and 

 it would be heard. 



If a thumb-screw with a scale be fitted on the tube through which 

 the stronger sound passes, by compressing the tube by means of the 

 screw the intensity of the sound can be graduated till nothing is heard 

 by the observer, thus proving the two to be equal, 



Reading the number of turns from the scale, the relative intensities 

 of the two sounds can be measured, and if the scale be made so that a 

 certain number of turns equals a sound of a certain intensity, the 

 measurement becomes positive instead of relative. 



