236 



PROCEEDINGS OF THE AMERICAN ACADEMY 



A large number of measurements were made with an inverted arc, 

 in which the jiositive carbon was below and the negative one above. 



Currents of approximately 5, 7, and 10 amperes were used, with 

 lengths of arc up to ^^ inch. There was much unsteadiness in the 

 voltmeter for arcs moi"e than ^.j inch long, there being a slow and 

 irregular fluctuation of the needle between definite limits. Very 

 probably this was due to irregularities in the air currents rising from 

 the lower and hotter positive carbon. In some sets of the measure- 

 ments (Series II. of tables) the extreme readings were taken and sepa- 

 rate curves drawn for each. The following tables (XI. to XVIII.) 

 give the results of the measurements. 



Inverted Arc. 

 TABLE XI. 

 Current = 5.1 amperes. 



Series I. , 



TABLE XII. 

 Current = 7.3 amperes. 



Inverted Arc. 

 TABLE XIII. 

 Current = 5.1 amperes. 



Series II. 



TABLE XIV. 

 Current = 10.2 amperes. 



