Nipher — The Nature of Electrical Discharge. 85 



dition bordering closely on fusion. Some parts were 

 fused. Fragments of the wire are shown in Plate IX, 

 Fig. A. At fairly regular intervals of 0.05 cm., the wire 

 was longitudinally compressed into nodules, shown in 

 the figure. This result may be due to electrical striae 

 within the wire. The wire cooled down to a rigid condi- 

 tion at the instant when it had assumed this form. The 

 nodules would then represent nodes of compression, and 

 the points midway between the nodules would be rare- 

 faction nodes in the electrical waves. Adjoining semi- 

 waves have alternately a surface super charge and a de- 

 ficiency of negative corpuscles. These semi-waves attract 

 each other, and the metal yields for the moment. 



In one case the discharge was slightly gi'eater, so that 

 the nodules separated, and further change in form was 

 arrested at the instant when they had an almost per- 

 fectly spherical form. There were about 1,000 of these 

 minute spheres distributed over a length of 50 cm. Those 

 at the ends of the tube were somewhat larger than those 

 along the main body of the tube, and were slightly dis- 

 torted. In cases where the discharge was greater, the 

 fragments of the wire were distorted into irregular forms. 



A sample of the fuse wire which is in its natural form 

 is also shown in Fig. A. The effect shown by the other 

 fragments in this figure were due to a single discharge. 



In Fig. B is shown a few cuttings from a copper wire 

 having a diameter of 0.09 mm., through which about 50 

 discharges were sent. This wire gradually buckles into 

 irregTilar wave forms, as was first observed by Plante. 

 Two wires, three meters in length were suspended in a 

 horizontal position on thirty-one fine silk threads at reg- 

 ular intervals. The threads were aliout two meters in 

 length. The wires terminated in somewhat larger wires 

 which show no such effects, which were bent downwards 

 into metal vessels containing a salt solution. A slight 

 expansion of the wire could be observed at each end when 

 a spark was sent through the wire. No recoil of the 

 wire as a whole could be observed, and summation effects 



