1905-6.] Electric Oscillations ancl Magnetic Properties of Iron. 41 
“ lag ” is equally well exhibited by both ; but for present purposes 
it is in my opinion also essential to state all the facts in terms of 
permeability (at cyclic extremes), of retentivity (when H = 0), and 
of coercive force (when B = 0), as has been done above. 
Magnetic Detectors of Electric Waves. 
The above experimental results have, in addition to their purely 
theoretical or physical aspect, an especial interest at the present 
time, in view of the fact that the magnetic properties of iron and 
steel have beeti utilised to detect electric waves in space. 
Magnetic detectors of the Rutherford type, either in their 
original forms or as they have been modified for continuous 
telegraphic work, illustrate what takes place when oscillations 
are superposed at one point of the normal hysteresis loop, viz., that 
where zero field has just been passed. The reduction of the 
magnetisation which occurs is sufficiently well understood. 
Other forms of detectors in which the adjustment and rates of 
motion of parts have been experimentally determined, seem to be 
less perfectly understood. If one may judge from the conflict of 
opinion which has arisen, and the anomalous results which have 
been obtained, magnetic detectors seem to have outrun very 
generally accepted theoretical knowledge. 
In view of the experimental results above arrived at, I propose 
now to discuss three forms of magnetic detectors of electric waves 
in the following order, viz., Marconi’s first form of detector, 
Marconi’s second detector, and the Ewing- Walter detector. 
First. In Marconi’s first apparatus a fixed core of iron or steel 
is used, long relatively to its diameter. The poles of a revolving 
horse-shoe magnet are, during each complete revolution, twice in 
close proximity to the ends of the core. It is thus continuously 
carried through a cycle between positive and negative maximum 
values. Wireless telegraphic signals are received at all stages of 
this cycle. The instrument is admittedly much less perfect than 
Marconi’s second form, but it is nevertheless interesting to examine 
its action. 
The broken curve of fig. 6 shows the effect of superposing 
and withdrawing oscillations at twenty-four approximately equi- 
