K L. Nichols — Alternating Electric Arc. 9 



Mr. Caldwell's first step, after having repeated the prelimi- 

 nary experiments of Archbold and Teeple, and verified their 

 statements, was to substitute a ball with a surface of platinum 

 for the brass ball used by them. The new ball withstood the 

 action of the arc no better than the old one had done. It 

 soon became covered with a black deposit, the growth of which 

 modified and vitiated the action of the apparatus. In experi- 

 menting with such a ball, the surface of which was still bright 

 and new, and with a point of the same metal, it was noticed 

 that within the critical distance, while the spark was passing 

 in both directions, there appeared to be two distinct paths 

 along which the discharge was taking place. One of these was 

 nearly in the line from the point to the ball, normal to the 

 surface of the latter, the other from the point in a direction 

 approximately at 45° with the common axis of the pointed rod 

 and ball. Upon increasing the distance until the discharge 

 entered the " one way " stage, the longer and oblique path of 

 flow vanished. In the revolving mirror the two classes of 

 sparks were easily indentified. They were found to occur in 

 alternation with each other, the spark which followed the nor- 

 mal path being that which passed from ball to point, the other 

 that from point to ball. The images of the discharge from the 

 point disappeared as soon as the critical distance was reached. 

 In order to place the matter beyond all doubt, the times of the 

 discharge which followed the normal path were determined by 

 an ingenious method, quite independent of that used by the 

 first observers, and it was found that the spark occurred always 

 in that part of the cycle during which the ball was positive. 



Mr. Caldwell's method of fixing the time of the discharge, 

 briefly stated, was as follows. An adjustable contact device, 

 similar to that used by Archbold and Teeple, was attached to 

 the shaft of the dynamo. A wire from one pole of a large 

 Holtz machine, driven by power, was carried to the neighbor- 

 hood of the ball and point, where two platinum terminals, I mm 

 apart were set up. The wire was connected with one of these 

 and a line was carried from the other to the contact device. 

 A wire from the latter to the remaining terminal of the Holtz 

 machine completed the circuit. Whenever the brush made 

 contact, a spark leaped between the platinum terminals, just 

 described. By adjustment of the brush, the spark could be 

 made to appear at any desired instant in the cycle of alterna- 

 tions of the dynamo. The platinum terminals were placed so 

 that the image of the spark in the revolving mirror was seen 

 side by side with that of the discharge between the ball and 

 point, and the precise position in the cycle, occupied by the 

 latter, was thus readily determined. 



Closer study of the two paths of discharge showed that the 

 oblique arc left the very apex of the point, swinging out later- 



