84 Trans. Acad. Sci. of St. Louis. 



The field thus impressed upon the wire was deficient 

 in its electrostatic component, necessary to produce an 

 apparent recoil of the wire. 



Under the conditions in which the creeping of the wire 

 occurs, we must consider the simultaneous discharge at 

 the two ends of the wire to be compression and rare- 

 faction waves traveling in opposite directions through 

 the wire. There was no visible expansion when the wire 

 was observed through the telescope. The i^henomena of 

 the cathode discharge discovered by Crookes, and the 

 fact that the plates of the influence machine are in con- 

 tinuous motion, give standing evidence of sudden dis- 

 placements at each spark discharge. The corpuscular 

 displacements in the compression and in the rarefaction 

 waves are in the same direction, namely, in the opposite 

 direction from that in which the wire creeps. It would 

 seem that such waves impressed upon the two ends of a 

 wire might result in the formation of a series of waves 

 throughout the entire wire. The corpuscular nebula 

 within an insulated sphere of metal is displaced when a 

 charged sphere is brought near it. It seems evident that 

 the conditions which exist on the two halves of such a 

 spherical surface may exist along a metal wire. They 

 would somewhat resemble longitudinal waves which might 

 be produced in a column of water contained in a flexible 

 i*ubber tube. The two halves of such an electric wave 

 would attract each other. 



A result which appears to give evidence of the existence 

 of such waves was obtained as follows: 



A piece of % ampere fuse wire, having a diameter of 

 0.115 mm. and a length of about 50 cm. was placed in a 

 glass tube containing coal-oil. The ends of the fuse wire 

 were soldered to copper wire of about 0.2 mm. diameter 

 which protruded through sealing wax with which the 

 ends of the tube were closed. A single discharge from 

 the machine to which was connected a condenser of about 

 4,000 sq. cm. area, gave various results depending upon 

 the spark-length. In two cases the wire reached the con- 



