Nipher — Properties of Light-struck Photographic Plates. 155 



While recently observing a plate exposed to the negative 

 discharge, half of which was shaded, a bright ball of light 

 looking like a globule of molten metal rolled slowly out into 

 the shadow from the brass weight. Its size appeared to be 

 that of a pin's head. It moved somewhat irregularly and left 

 a black narrow track in its wake. The discharges which had 

 previously been in all directions around the medal were now 

 all on the side occupied by the ball. They were less frequent, 

 and the sparks were apparently within an angle of 30° or 40°. 

 They, however, did not pass along the track of the ball, but 

 rather appeared to avoid it. Another ball appeared from the 

 same point on the coin and the first disappeared. The second 

 ball soon diverged from the track of the first, and slowly made 

 its way outwards. Several others followed. The plate was 

 finally developed, and these tracks appeared as a branching 

 system of black lines, wholly unlike anything before observed. 

 The experiment was repeated, and a similar result followed. 

 It was then thought that the shadow in which they made their 

 appearance might be concerned in the phenomenon. In the 

 next plate, however, the balls appeared in the part of the 

 plate which was strongly illuminated. The development of 

 this plate was pushed to the extreme, and the branching- 

 track began to appear blurred. On examination with a pocket 

 lens, it was found that the tracks of the ordinary spark dis- 

 charges could be detected where they crossed the tracks of 

 the ball discharges. They were here intensified. Under the 

 glass the spark discharges appeared indistinct and hazy, while 

 the tracks of the balls were still sharp. See Fig. 1. When 

 rephotographed and enlarged 100 diameters, the tracks of the 

 balls on the original negative were found to be about 0.002 cm. 

 in width. 



Some changes were then made in the apparatus in order to 

 provide more suitable conditions, and it was then found im- 

 possible to secure the result again after two days of persistent 

 work. The old arrangement was then resorted to with like 

 results. It was observed that the ball discharges came from 

 the same point on the brass weight. A short radial pencil 

 line was drawn on the plate at this point with no result. 

 There seemed to be no irregularity at the point of discharge 



