May 17, 1895.] 



SCIENCE. 



535 



Figure 2 shows the result of a simihir 

 series with the magnetizing foree for curve 

 1 greater and the operations 2 and 3 of 

 figure 1 omitted. Figure 3 illustrates the 

 result when the w-hole of the reverse cur- 

 rent was put on in operation 1, and the 

 curve 2 shows the effect of short circuiting 

 the hatterj' in the second cii'cuit. Figure 

 4 is the same as figure 3 so far as the first 

 operation is concerned, but in the operation 

 which gave curve 2 the second magnetizing 

 circuit was simplj' broken. The scales of 

 these curves are arbitrary, but are the same 



complete bi-eak of the magnetizing coil 

 circuit. 



The fact that the time required to pro- 

 duce the change of magnetization is depend- 

 ent on the amount of change shows that, 

 unless the period of the galvanometer 

 needle be so long that even the longest of 

 these times is short in comparison, the 

 measurements of the higher magnetizations 

 will be more in error than the lower. The 

 effect of this on the magnetization curve of 

 iron is to render the steep parts of the curve 

 less steep. The curves 1 and 2 of figure 4 



ES 



for the different cui-^-es, and hence the rela- 

 tive magnitudes of the changes of current 

 may be estimated from the curves. The 

 reverse current in the second coil was not 

 at any time adjusted so as to give an equal 

 but opposite magetization to that given by 

 the coil through which the constant current 

 was kept flowing. The two primary objects 

 of drawing the curves were (a) to show 

 the great difference in the time required to 

 produce changes of magnetization as de- 

 pending on tlie magnitude of the change, 

 and (b) to show the differences in time 

 for the two cases of short circuit and 



show the efiect of tlie diminished inductive 

 retardation when the circuit is broken in 

 shortening the time required for the mag- 

 netization to change back as compared with 

 the time required to produce it. Curve 2 

 of figure 3 compared with curve 2 of figure 

 4 shows the relative times when in the first 

 case the e. m. f. is removed, but the circuit 

 left closed and in the other case the circuit 

 is broken. Comparisons between the de- 

 flection due to the application and the re- 

 moval of magnetizing force should always 

 be made in such a way that the circuit has 

 the same inductive retardation in both cases. 



