OF ARTS AND SCIENCES. 339 



from a moderate value, in parts of the bar not too remote, to infinity, 

 somewhere in its central portion, and then descends with change of 

 sign, for a still further advance in distance, to a minimum, and finally 

 becomes infinite again ; so that, the most distant portions of the wave 

 reaching their maximum height simultaneously with the nearest, the 

 wave appears to flow backward over the most distant part of the bar. 



VIII. The apparent velocity of propagation over the first fourteen 

 diameters is from twenty-five feet per second, corresponding to four 

 reversals per second, up to ninety feet, for twenty-three reversals, after 

 which it rises nearly in proportion to the number of reversals, until, for 

 one hundred and forty reversals, it measured about three hundred feet 

 per second. 



IX. The magnetic conductivity of the bar increased (in one case 

 from two to three fold) under the continued influence of the alternat- 

 ing current sent by a high electromotive force, wdiose use in these 

 determinations must generally be avoided. 



X. The retardation of phase is less for small rods than for large 

 ones, but in most respects the effects are similar for distances propor- 

 tional to the relative diameters of the bars. 



XI. The retardation is less for steel than for iron. With steel, a 

 high electromotive force may be adyantageously employed to over- 

 come the tendency of the magnetic displacements to become per- 

 manent. 



XII. The phase of the magnetic wave may be treated as the result- 

 ant of that of two sets of waves, one due to the direct (and instan- 

 taneous) action of the primary itself, and proportional, accordingly, 

 to the time and to the inverse cube of the distance ; the other depend- 

 ing upon the secondary .effect of magnetism already induced in the 

 bar, and hence proportional more or less to the square and higher 

 powers of the time, and to the distance as the exponent of a constant 

 factor, less than unity. In the same way that each law of magnetic 

 induction has a parallel in the conduction of electricity, it is probable 

 that the law of change in the one is similar to the law of change in 

 the other. 



