July i6, 1903] 



NATURE 



249 



pressed and the circuit closed ; the vibrating arm opens 

 and closes the circuit of the second sounder, to which 

 is attached the dipping rod of the mercury break. It 

 is said that this arrangement gives a more regular 

 succession of sparks than is obtained with one sounder 

 only. An automatic transmitting apparatus has also 

 been worked out by Messrs. Lodge and Muirhead. 

 This is shown at the right of Fig. i, in front of the 

 buzzer, and consists of two pieces of apparatus, a per- 

 forator and a transmitter, which are used in conjunc- 

 tion with the buzzer, &c., in place of the ordinary sig- 

 nalling key. 



A regular succession of sparks having been thus 

 obtained, still only part, and that the simpler part, 

 of the difficulty has been overcome, for it is not the 

 period of the sparks but the period of the oscillations 

 in the spark which has to be syntonised. When one 

 considers how short is the train of waves from each 

 individual spark and how long comparatively the 

 interval between two successive sparks, it is easy to 

 see the importance of getting the best results possible 



! detail in Figs. 2 and 3. Fig. 2 represents the com- 

 ! plete receiving instrument. The instrument looks at 

 I first sight much like a Morse recorder; the coherer is 

 mounted behind the box which contains the clockwork 

 i for feeding forward the tape and rotating the coherer 

 wheel. Its construction (fan be seen from Fig. 3, which 

 shows a coherer by itself. It consists of a small steel 

 disc with a fine razor edge which dips into a little pool 

 of mercury in an ebonite cup. The mercury is covered 

 I by a thin film of oil, and the disc is adjusted so that 

 I under normal conditions the oil serves just to insulate 

 it from the mercury. When oscillations are set up in 

 the coherer circuit, this thin layer of insulation is 

 ; broken down, and connection established between the 

 disc and the mercury. The disc is slowly rotated by 

 means of the notched wheel seen clearly in the illus- 

 tration, which gears with a similar wheel at the back 

 i of the clockwork box. Connection is thus no sooner 

 established between the disc and mercury than it is 

 broken again by a fresh oily portion of the edge coming 

 round; there is consequently only connection during 

 the time the oscillations are actually arriving and the 



from each spark. Herein, indeed, seems to lie one of 

 the chief unsolved problems of wireless telegraphy — 

 the problem of obtaining a really continuous series of 

 undamped oscillations. It seems doubtful whether, 

 even with the best possible design and arrangement of 

 apparatus, a satisfactory solution will ever be found 

 by means of disruptive sparks. Perhaps we must look 

 to some quite different method of setting up the 

 oscillations. The method that gives most promise of 

 ultimate success is some application of the principle 

 of Mr. Duddell's musical arc, as suggested by Mr. 

 Duddell at the Royal Institution last year (see also 

 the Electrician, May i, vol. li. p. 84). It certainly 

 seems that from this discovery may be developed a 

 means of producing a continuous series of undamped 

 oscillations of high frequency, and if this should prove 

 to be possible a change amounting almost to a revolu- 

 tion would be effected in the practice of syntonic wire- 

 less telegraphy. 



We may pass now to a consideration of the receiving 

 instruments which are shown in Fig. i, and in more 



NO. 1759, VOL. 68] 



Fig. 3.- The Coherer. 



[coherer is self-decohering and requires no tapping 

 'back. In some respects the device recalls a sugges- 

 tion made by Rupp five or six years ago, who proposed 

 mounting a filings coherer so that it was rotated slowly 

 by the Morse tape. The Lodge coherer is, however, a 

 far more mechanical contrivance than a filings tube 

 however the latter may be decohered. In order to keep 

 the edge of the disc clean a pad of felt is pressed lightly 

 against it; this can just be seen on the left near the 

 top of the disc; contact is made by a spring pressing 

 against the shaft on which the disc is mounted. The 

 coherer will only work with a very small potential dif- 

 ference — a fraction of a volt — between mercury and 

 disc; it is therefore connected in series with a potentio- 

 meter, which reduces the voltage from the cell. 



Another feature of the receiving circuit is the absence 

 of any relay; the coherer and potentiometer are 

 directly in series with the recording instrument, which 

 takes the form of a simple syphon recorder. This is 

 seen on the right of the clockwork in Fig. 2 ; the pert 

 consists of a fine glass syphon tube suspended from 

 the galvanometer coil, one end dipping in a cup of ink 



