114 DR. W. G. DUFFIELD, MESSRS. THOS. H. BURNHAM AND A. H. DAVIS ON 



Method 4 is a good one save for the effects of convexion currents which are 

 considerable, because they rise over the curved side of the carbon. 



TABLE III. Couples Due to a Single Arc. 



Regarding the experiments upon alternating current arcs as a whole the results 

 (Table I.) show a satisfactory measure of concordance for the various methods. 



y 

 I 



g - 

 I-O- 



ri 



<o 



0-S- 



Alternating 

 current 



-l 1 r 1 1 1 f I r- 



5 loAmps. 



Fig. 6. From Table I. 

 Method 1 . . . O- Other methods . 



Fig. 6 shows the results graphically, open circles representing the observations 

 obtained by method 1, full circles the results of the other three methods. 



