174: Perkins — Velocity of the Propagation of Magnetism. 



245 ft / sec as against 257 ft / sec actually observed. It must be 

 admitted that values of v calculated for other points along the 

 bar did not agree so closely with observation, those nearer the 

 center being too small and those farther out too large ; but 

 considering the fact that equation (5) makes 7 exceedingly 

 sensitive to very small errors in the values /3, C and w, a closer 

 agreement could hardly be expected. 



A more approximate method suggested by Zennech gives 

 &ly = 2*4 for cases of moderate permeability and frequencies of 

 about fifty cycles. Using /S = *16 this gives v = 5712 cm /sec> 

 which is as close as could be expected from the approximate 

 nature of the assumptions. 



In conclusion we may, I think, accept as proved that the 

 velocity of magnetism in metallic paths, and especially those of 

 high permeability, is small compared to that in non-conducting 

 bodies. That this effect is due to eddy currents and therefore 

 variable, diminishing as the flux diminishes along the bar. 

 That this velocity is not a simple function of a?, but depends on 

 the physical properties of the bar, such as ohmic resistance and 

 permeability, as well as the magnetic density at the point in 

 question. And that this retardation is great enough to cause 

 a very perceptible lag where the density is high and the lamina- 

 tion poor; great enough indeed, it would seem, to call for 

 recognition in the design of many forms of electromagnetic 

 machinery. 



Trinity College, July, 1904. 



