1244 THE BELL SYSTEM TECHNICAL JOURNAL SEPTEMBER 1953 



for both active and clean contacts the transient started with a rapid 

 drop. For clean contacts this drop was directly to the steady arc voltage 

 of the contact metal. For active contacts the drop was to a higher voltage 

 between 15 and 32 volts followed by a gradual and irregular drop in 

 voltage. The time of the first rapid drop was invariably between 2 X 

 10"^ and 4 X 10"^ second for both clean and active contacts over a range 

 of circuit inductances between 0.1 X 10~^ and 48 X 10~^ henry. This 

 time was just about the time resolution of the scope used indicating that 

 in all cases the initiation time was less than the time resolution of the 

 scope. This was also borne out by our calculations, Equation 3, where 

 the longest initiation time for the conditions studied was only 1.2 X 

 10~^ second. The higher arcing voltage of active contacts mentioned 

 above has been previously reported in Reference 3. It was pointed out 

 that active contacts have arc voltages comparable with those of carbon, 

 19 to 30 volts.* The slow drop in voltage following the first rapid drop 

 observed with active contacts is probably a burning off process of the 

 activating substance on the contacts as indicated by a continuous ap- 

 proach of the arc voltage to that of the clean metal. It may be added 

 that for cases where the initial circuit voltage is closer to the carbon 

 arc voltage one should expect a smaller initial drop. This was confirmed 

 by measurements made at 35 volts. 



In connection with the experimental study of the initiation of the 

 arc, the following concluding remark may be made. Unless the measuring 

 apparatus has a response faster than the individual discharges at the 

 contacts, the recorded transient will essentially be some particular 

 average of a complex contacts transient. It should not, accordingly, 

 be mistaken for the more fundamental and usually much faster formative 

 transient of the arc. 



The author is indebted to Dr. P. Kisliuk and Dr. L. H. Germer for 

 much valuable discussion. 



* Recent measurements by the writer on arc lamp carbon have given arc volt- 

 ages as high as 43 volts. 



