204 Mr. G. Marconi [June 13, 



enumerated when worked in connection with my system of syntonic 

 wireless telegraphy. 



These coherers are, however, useful if employed for temporary tests 

 in which the complete accuracy of messages is not all- important, and 

 when the attainment of syntonic effects is not aimed at. They are 

 especially useful when using receiving vertical wires supported by 

 kites or balloons, the variations of the height of the wires (and, there- 

 fore, of their capacity) caused by the wind making it extremely 

 difficult to obtain good results on a syntonic receiver. 



Coherers have long been considered as constituting almost the 

 essential basis of electric space telegraphy, and although many other 

 detectors of electric waves existed, none of them possessed a sensi- 

 tiveness which even approached that of a coherer, and most of them 

 were also unsuitable for the reception of telegraphic messages. With 

 a view to producing a receiver which could be worked at a much 

 higher speed than a coherer, I was fortunate enough to succeed in 

 constructing a magnetic detector of electric waves, based on a prin- 

 ciple essentially different from that of the coherer, and which I think 

 leaves all coherers far behind in speed, facility of adjustment and 

 efficiency when worked in tuned circuits. This detector, which I 

 had the honour of describing in detail before the Royal Society yes- 

 terday, possesses I believe a sensitiveness which surpasses that of the 

 best coherers. The magnetisation and demagnetisation of steel 

 needles by the effect of electrical oscillations has long been known 

 and was noted especially by Prof. J. Henry, Aloria, Lord Kayleigh 

 and others. Mr. E. Rutherford also has described a magnetic detec- 

 tor of electric waves based on the partial demagnetisation of a small 

 core composed of fine steel needles previously magnetised to satura- 

 tion. By means of a magnetometer Mr. Rutherford succeeded in 

 1895 in tracing the effects of his electrical radiator up to a distance of 

 three-quarters of a mile across Cambridge. But Mr. Rutherford's 

 arrangement is not suitable for the reception of telegraphic messages 

 in consequence of the fact that a careful process of re-magnetisation, 

 which requires some time to effect, is necessary in order to restore 

 its sensitiveness after the receipt of each impulse. Mr. Rutherford's 

 arrangement is also considerably less sensitive than a coherer. 



The detector which I am about to describe is, in my opinion, 

 based upon the decrease of magnetic hysteresis, which takes place* in 

 iron when under certain conditions it is exposed to the effect of high 

 frequency oscillations or Hertzian waves. As employed by me, it 

 has been constructed in the following manner : — On a core of thin 

 iron or steel, but preferably hard drawn iron, are wound one or two 

 layers of thin insulated copper wire. Over this winding, insulating 

 material is placed, and over this, again, another longer winding of 

 thin copper wire contained in a narrow bobbin. The ends of the 

 windings nearest the iron core are connected, one to earth and 

 the other to an elevated conductor, or they may be connected to the 

 secondary of a suitable receiving transformer or intensifying coil, 



