328 PROCEEDINGS OF THE AMERICAN ACADEMY 



The apparent velocity of propagation over the first seven inches of 

 the bar ranged from 25 feet per second, for the low velocity of four 

 reversals per second, up to 90 feet for about 23 reversals; after which 

 it rose nearly in proportion to the number of reversals, until for 140 

 reversals it measured about 300 feet per second. At distances greater 

 than 7 inches, the direct and instantaneous action of the primary and 

 adjacent parts of the bar predominates more and more over the pulse 

 propagated through the medium of the iron ; so that the investigation 

 of the rate of propagation in distant portions of the bar is not in- 

 structive. Indeed, since the most remote portions of the bar agree 

 in phase with the primary, while intermediate portions lag behind it, 

 we shall find the magnetic wave apparently flowing backward in 

 the most distant portions. This, however' paradoxical it may seem, is 

 strictly in accordance with both theory and fact. 



The following set of observations was dated Jan. 11 : — 



c =■ the arbitrary position of the coil. 



ph =■ the position of the sliding-piece which gave zero deflection 

 of the galvanometer. 



T^= the number of ticks of a watch (beating 288 times to a min- 

 ute) made while the shaft was performing 72 complete revolutions, 

 and counted before and after each set of observations. 



7=84 - - - 83. 7=25 - - - 29. 



The first point of simultaneous action was 10^ mm., the second 

 1^\ mm., and the third 38* mm. The distances were reckoned in 

 inches, beginning at a point 2 inches from the centre of the primary. 



Hence we get the following tables, where v = the number of revo- 

 lutions per second, t = the length of time occupied by the wave in 

 reaching the distance, d, measured in inches from the centre of the 

 primary : — 



