ST. JOHN. — WAVE LENGTHS OF ELECTRICITY. 225 



Thompson has shown that iron absorbs more energy than brass when 

 subjected to rapidly alternating magnetizing forces, while Trowbridge 

 and Bjerknes agree in showing that iron and nickle conductors damp 

 out electric waves much more rapidly then copper, and all agree that 

 the magnetic properties of iron are called into play under such 

 conditions. 



It has seemed to the writer that it was still an interesting and 

 profitable field to investigate whether the magnetic properties of 

 iron were acted upon sufficiently, and in such a way as to produce a 

 change in self-induction that would affect the rate of propagation 

 of electric waves along iron wires. With this end in view, it was 

 sought to simplify some of the apparatus hitherto employed, and to 

 arrange it so that the effect, if observed, could not be due to any other 

 cause than the magnetic properties of the metals experimented upon. 



Description of the Apparatus. 



For producing the oscillations, the ordinary Hertz vibrator (Fig. 3) 

 was employed, consisting of two plates of zinc, each 40 cm. square, 

 61 cm. apart, and mounted upon insulated wooden supports. These 

 supports had square wooden bases that could slide in a grooved plank, 

 so that their distance apart could be easily varied, and by turning the 

 supports through a quarter revolution the plates could be made to face 

 each other or to stand m the same vertical plane. The plates were 

 joined by a conductor of brass 0.5 cm. in diameter, which was connected 

 to the plates by sliding with friction into* brass tubes 15 cm. long sol- 

 dered to the plates. The conductinij wire was broken by a spark gap, 

 provided with brass balls 3 cm. in diameter. Tin balls were tried, and 

 had perhaps a little greater effect in exciting oscillations in the secon- 

 dary ; but they required polishing fully as often as the brass balls. 

 Finally brass balls with platinum faces were used, and were found to 

 be much more constant in effect than either brass or tin. A circular 

 piece of platinum 0.025 cm. thick and 2 cm. in diameter was wedged 

 into a shape to fit the front of the ball and there countersunk so that the 

 joint was smooth. At first a piece of platinum 1 cm. in diameter was 

 tried, but this was too small, and frequently the sparks jumped from 

 the edge of the platinum or brass. These were discarded, and the 

 larger platinum faces employed. In a very few minutes after polishing 

 they would often reach a state that would remain constant through a 

 long series of observations, but frequently several trials would have to 

 be made before such a satisfactory condition was obtained. All the 



VOL. XXX. (n. S. XXII.) 15 



