234 



PROCEEDINGS OF THE AMERICAN ACADEMY 



quality, and the intensity of the sound transmitted, were noted. The 

 accompanying plots (Figure 3) and table will serve as exanqdes of 

 the series. 



The values given are those of the steady currents, which, in the 

 absence of any magnetic effects from the earth, would give the de- 

 flections observed. As the precise form of the variation of the current 

 is uncertain, no better mode of procedure has suggested itself. The 

 columns headed Def., C, and Wt., contain the observed deflections, 

 the calculated currents, and the weights in the pan. These weights 

 should be divided by 2 to obtain the pressure at the contact. 



BRUSH STORAGE BATTERY, TWO CELLS IN PARALLEL. 



It will be noticed that the current rises very rapidly at first with 

 this increase of pressure. At all points of this rapid rise the sound 

 transmitted is very bad, and there are very frequent breaks with the 

 intensity of sound employed. The maximum is soon reached, at 

 about 1,000 mgr. pressure, and from that point the current falls off 

 gradually. The sound becomes good soon after the maximum cur- 

 rent is reached, and as the pressure increases the sound diminishes 

 in intensity but improves in quality. In all our experiments the 

 same form of curve represented the variation of pressure and current, 

 and in all, the best sound was transmitted directly after the maximum 

 current. 



I find that this form of curve is in harmony with theory, assuming 

 that the pressure at a microphone contact varies inversely as its 

 resistance. The total resistance in the circuit is a constant plus the 

 resistance of the joint. This may be expressed by the equation 



/3 



i? = a + 



P> 



