178 THE ROYAL SOCIETY OF CANADA 



chinery of the power house was clearly heard. This was probably 

 caused by the slightly changing voltages due to the peaks of the 

 waves in the direct current supplied. 



It is rather fascinating to listen in this way to what is taking 

 place inside the lamp, the thermionic current varying in sympathy 

 with the power house. Sometimes the rushing or hissing noise would 

 reappear, perhaps due to "strays" entering the Physics Building 

 along the hot water pipes. These strays probably come, as Dr. 

 Barnes suggested to me, from the street railway system, some half 

 mile away, and are mainly dominant when a street car starts from 

 rest. 



An arrangement of the kind described might conceivably be used 

 for detecting "strays". 



15 Feb. 1915. 



With a direct current from a dynamo there is sufficient variation 

 in the voltage to obtain a sound just audible at 100, loud at 110, 

 very loud at 130, and it might be described as an uproar at 140 volts. 

 It might be expected that the intensity would increase until the 

 lamp burnt out. Nothing of the sort. At 142 volts, the uproar is 

 replaced by dead silence which continues up to 165 volts, as high 

 as the lamp would stand. 



The explanation may be gathered from Langmuir's paper, (Physic- 

 al Review, Dec, 1913). The thermionic current does not increase 

 with the temperature according to Richardson's Law, unless the 

 vacuum is of a high order. With a moderate vacuum, the volume 

 charge between the filament and silver causes the thermionic current 

 to remain at a value nearly constant as the temperature is raised above 

 a certain value. The thermionic current begins by obeying Richard- 

 son's law (i = cdre — 6 ) and then later approximates to a steady value. 



Thus at low voltages, variations of voltage cause variations of 

 temperature and consequent fluctuations of thermionic current, 

 heard in the telephone. 



Above 142 volts, however, for the particular lamp in question, 

 a change of voltage and of temperature produces no change of current, 

 and hence no sound can be heard in the telephone. The change 

 from noise to silence is remarkably abrupt. 



It is possible that this method may prove very convenient for 

 testing the electric emission from various sustances, in different gases, 

 and it suggests a method of measuring a high vacuum. 



