PIERCE. — A METHOD OF MEASURING THE INTENSITY OF SOUND. 379 



a maximum deflection. The adjustment of the contact is made once 

 for all, and subsequent accidental changes of the apparatus is prevented 

 by filling the cavity about H with melted wax or plaster of Paris. 



When made in this manner the rectifier will stand considerable 

 abuse in the way of jar and overload. It is, however, subject to 

 changes due to the expansion and contraction of the mounting, and 

 due also possibly to a temperature coefficient of the molybdenite itself. 

 Effort to get a mounting without such changes with temperature and 

 a study of the temperature coefficient of the substance itself are now 

 in progress. Up to the present it is found advisable to use the rectifier 

 in a thermostat at constant temperature, when accurate quantitative 

 agreement between observations extending over a considerable period 

 of time is required. 



Whether or not the direct current obtained from the molybdenite 

 in contact with two unequal electrodes is a thermo-electric action due 

 to the unequal heating of the electrodes by the oscillating current is 

 at present not known. It will be seen that the conditions are favor- 

 able for such thermo-electric action. In order not to commit one's 

 self to any particular theory as to the nature of the action, the device 

 is here referred to as a "rectifier," in that the current in one direction 

 due to an impressed voltage is very different from the current in the 

 opposite direction under the same voltage. 



III. Electric Circuits Employed with the Molybdenite 

 Rectifier in Experiments on Sound. 



In the measurement of sound, the rectifier was at first placed 

 directly in series with a sensitive galvanometer and a Bell magneto- 

 telephone receiver. With this arrangement, when sound was made in the 

 neighborhood of the receiver, the vibration of the telephone diaphragm 

 generated electric oscillations in the circuit. These oscillations passed 

 through the rectifier more strongly in one direction than in the oppo- 

 site direction, and caused a deflection of the galvanometer. 



However, on account of the high resistance of the rectifier, and in 

 order to take advantage of electrical resonance in the circuits, it was 

 found better to employ an arrangement of circuits containing a 

 step- up transformer, as is shown in Figure 2. 



In Figure 2 PS is a transformer, the primary P of which is con- 

 nected in series with the telephone T and an adjustable condenser C. 

 The secondary S of the transformer is connected in series with the 

 rectifier R, the galvanometer G, and a calibrating device at W. By 

 adjusting the condenser C, the electric circuit TCP was brought to 



