526 



PROFESSOR KNOTT ON SOME RELATIONS BETWEEN 



much diminishes the area of the cyclic graph. The symmetrical character of the curve 

 for nickel as shown in fig. 5 is very remarkable, especially since in the case of iron 

 (fig. C) there is distinct asymmetry. Here we have an iron wire twisted through an angle 

 of 90° and subjected to a variation of current from — 3 '94 amperes to + 3'4 and back to 

 -3 - 47. The original negative polarity makes itself felt throughout the whole cycle, so 

 that the positive longitudinal intensity never exceeds 150, although the greatest negative 

 intensity obtained was 260. As is very distinctly brought out in fig. 6, the negative 

 intensity first increases as the current is diminished, and begins to diminish only after 

 the current has been changed in direction, and has attained a value of about half an 

 ampere. This is a very remarkable result, and hints at a very complex magnetic 

 aeolotropy. 



Another peculiarity in connection with the reversal of the current is worthy of notice. 

 When the wire was untwisted from either of its limiting positions to some intermediate 

 stage, and the current then reversed, the intensity changed in a manner quite different 

 from what was observed when the wire was fully twisted. Take, for example, the follow- 

 ing experiment with nickel wire, with which, at each successive stage of twist as 

 indicated in the first column, the current was varied from its positive value to an equal 

 negative value, and back again to the original positive value. After the completion of 

 this cycle, the twisting was continued through its complete cycle and one stage more. 

 At this stage the current was again varied cyclically, then the twisting continued through 

 another cycle and one stage more, and so on till throughout one half cycle of twisting the 

 effect of reversing the current at every stage had been studied. The second, third, and 

 fourth columns contain the intensities corresponding to the limiting values of the 

 cyclically varied current ; the fifth column gives the range, and the sixth the mean 

 polarity. 



;;=; ^«^ 



^%^ 



+ 3-4 



-3-4 



+ 3-4 



Range. 



Mean. 



-225 



+ 232 



-186 



+ 222 



+ 413 



+ 21 



-180 



+ 222 



-172 



+ 210 



+ 388 



+ 22 



-135 



+ 188 



+ 98 



+ 158 



+ 271 



+ 38 



- 90 



+ 121 



+ 73 



+ 100 



+ 48 



+ 92 



- 45 



- 33 



+ 100 



- 22 



-128 



+ 36 







-134 



+ 134 



-122 



-162 



+ 3 



+ 45 



-190 



+ 174 



-182 



-360 



- 6 



+ 90 



-212 



+ 193 



-202 



-400 



- 7 



+ 135 



-222 



+ 205 



-212 



-422 



- 6 



+ 225 



-233 



+ 213 



-222 



-441 



- 8 



In this particular case the reversal of the current at the stages of greatest twisting 

 does not produce quite so great an effect as the reversal of the twist for steady current. 

 There is nearer approximation, however, to a symmetrical change than in the case of 

 several of the other stages. Notably is there a lack of symmetry for the stage — 90°, in 

 which the polarity is throughout positive, the range of intensity being comparatively small. 



