154 Trans. Acad. Sci. of St. Louis. 



a lower potential than that of the halo around the larger 

 weight. With a spark length between thirteen and twenty 

 millimeters, disruptive sparks begin to appear around the 

 medal, and the discharge picture changes its character en- 

 tirely. Not only is it greatly different from those previously 

 described, but as in all other cases it does not correspond to 

 the appearance which it presents to the eye. See Fig. 5. 

 The visible spark discharges show a curious tendency to turn 

 at right angles, and seem to be unsteady and flickering in their 

 outer extremities. As the spark length increases the disrup- 

 tive discharges become several inches in length, and the gen- 

 eral appearance of the field as shown in the negative, is shown 

 in Fig. 6. In this figure the spark is delivered to the larger 

 disk. In most of these cases the development has been ar- 

 rested before the tracks of the visible sparks appeared on the 

 negative. No discharge like those shown on the negative can 

 be detected by the eye. If the knobs of the machine are 

 separated so widely that no sparks can pass, the brush dis- 

 charge gives very feeble results if exposure and developing 

 are otherwise done in the same way as before. This applies 

 both to the image of the object and to the field around it. 

 The oscillating spark discharge appears to be the important 

 element rather than luminous or electrolytic action. Certainly 

 the luminous effect is distinctly prejudicial. 



The negative discharge shows much less of interesting detail. 

 With short sparks, there is a smooth corona, looking like a 

 brush shading in India ink. With longer sparks some radial 

 line-work suggesting lines of force appears. See Fig. 7. 

 The general appearance is much the same for short as for 

 long spark lengths. When strongly illuminated during the 

 electrical action, both positive and negative discharges give 

 weak coronal effects in the negative, and the color is that of 

 a sepia stain, or an untoned silver print. Most of the in- 

 teresting features which the negatives show must be passed 

 over without mention. They pertain largely to effects due 

 to variation of capacity and spark gaps both in the main and 

 the parallel circuits. The perfecting of the methods and the 

 study of these features occupied a period of several months, 

 and a large part of this work was done during 1896. 



