648 MR JAMES RUSSELL ON 



in § 19, reversals of H 2 merely accentuate the very considerable molecular disturbance 

 caused by the superposition of the unidirectional force at right angles. It also appears 

 probable, judging from the readings given in the first, third, fifth and seventh sets of 

 determinations given in Table VIII. , that had the amplitude of the H 2 reversals been 

 less, the relatively large effect of the first superposition would have been somewhat 

 reduced in comparison with subsequent reversals. We know that the total ordinate 

 change of B x would have been reduced ; and, in any case, it is interesting to observe 

 that it is the first superposition, followed by a very few reversals of the superposed force, 

 which produces the total ordinate change of B r Further alternations of H 2 are powerless 

 to effect more than a partial collapse of the hysteresis loop due to the force first acting 

 when the amplitude of these alternations is insufficient. 



§ 28. The above results obtained with superposed forces slowly alternating may be 

 compared with similar effects due to high frequency discharges. Mr E. Rutherford,* 

 in his experiments on the demagnetising effect of such discharges on needles magnetised 

 to saturation, found that after a steady state was reached " the passage of further dis- 

 charges has no apparent effect on the magnetisation of the needle." He obtained the 

 further result, that the magnetometer deflection, measuring demagnetisation, " was 

 approximately proportional to the magnetic force acting upon the needle, provided the 

 magnetic force was well below the value required to completely demagnetise the 

 steel." 



Mr Rutherford was able to detect high frequency waves in free space at a distance 

 of over half a mile by means of a detector constructed on above principle, and recently 

 Signor Marconi t has made use of the extreme sensitiveness of that part of B-H 

 induction cycle where B is changing most rapidly with respect to H, as a detector of 

 electric waves for space telegraphy. 



§ 29. As the ordinate changes of B x above described reach their final values at the 

 extremes of the superposed reversals, the curves become closed upon themselves, and 

 further changes of the B 1 induction component are strictly cyclic. This final change of 

 the B x component falls to be co-ordinated with the cyclic change of the B 2 component 

 impressed upon the iron by the alternating superposed force H 2 . The cyclic values of 

 B 2 were measured under the same conditions as prevail in fig. xv. In this case B 2 

 corresponds with the circular magnetising force. It was found that the B 2 /H 2 cycle, 

 when the force first acting upon the iron was H ? = 20'4 C.G.S. units, differed some- 

 what from the same cycle when H z =0, — or, in other words, from the B-H cycle given 

 in fig. vii. for shield B. The maximum induction values at the extremes of the cycle 

 were reduced by about ten per cent. The positions on the cycle at which the values of 

 H 2 changed most rapidly with respect to H 2 were not materially altered, and occurred as 

 the horizonal axis was crossed, as in the normal cycle (fig. vii.). But the maximum 



* Phil. Trans., " A Magnetic Detector of Electric Waves," vol. clxxxix., 1897. 



t Paper read before Royal Society 12th June 1902, and entitled "A Magnetic Detector of Electric Waves which 

 can be employed as a Receiver for Space Telegraphy." 



