August i8, 1910J 



NATURE 



225 



extinguished, and anottier tap of the Morse key causes it 

 to light again. 



Now suppose that the taps of the Morse key were con- 

 trolled by the lines in a photograph or sketch, and that 

 tlie light from the lamp were concentrated on a revolving 

 photographic film, and you will see at once how a photo- 

 graph could be transmitted by wireless telegraphy. 



Such a process would be utterly impracticable com- 

 mercially, but my telectrographic system can be used with 

 success in its place. A line picture prepared in the way 

 already described is attached to the drum of the trans- 

 mitter, and the intermittent current, which is ordinarily 



eant'Ti 



passed into the telephone line, goes into an electromagnet, 

 M in Fig. 10, which then attracts a soft iron diaphragm 

 attached to brass springs, which are fi.xed to two rigid 

 supports. Every time current flows through the magnet 

 coils this diaphragm is attracted to it, and the platinum 

 contacts PQ are brought together ; when the current flows, 

 and PQ are in contact, the primary circuit of a trans- 

 former is closed, and the secondary having a spark gap 

 and being inductively coupled to the aerial and earth, a 

 signal is transmitted into space. Thus in the wireless 

 transmitter the only difference from ordinary telegraphy 

 lies in the fact that the length of the signals and their 

 distance apart are regulated by the lines composing the 

 sketch or photograph. 



When working with high voltages in the primary, such 

 as no, arcing is liable to take place, and hence the 

 distance between p and q when not attracted must be 

 considerable. This means that the distance between the 

 diaphragm clamps, 55 in the figure, must be short, and 

 the German-silver spring of which the diaphragm is made 

 must be thick, these two conditions making the natural 

 period of vibration very short. I have, however, found 

 that by interposing a mercury motor-interrupter in the 

 primary circuit, arcing is almost entirely avoided, as if an 

 arc be formed the current is interrupted an instant later, 

 and the arcing ceases in consequence. 



The receiving apparatus is very simple, and depends, for 

 short-distance work, upon a coherer cymoscope, the 

 decohering apparatus being of a particular character. 

 Every time an oscillation passes to the antenna, the coherer 

 becomes conductive in the ordinary way, and a relay is 

 actuated ; this relay is usually made to start a hammer 

 vibrating, the hammer hitting the coherer, and thus 

 causing it to lose its conductive power. But a vibrating 

 hammer is useless for the photo-telegraphic receiver, and 

 it is essential to have one strike only on the coherer for 

 each signal detected. 



The form of apparatus I have employed for this purpose 

 is seen diagrammatically in the next lantern-slide (Fig. ii). 

 EE is the magnet which is actuated by the relay R. It 

 then attracts an armature MN, which moves towards the 

 magnet poles and brings a resilient hammer H, fitted with 



NO. 2129, VOL. 84] 



a platinum contact p, against the coherer. The coherer 

 .\B is also fitted with a collar F and contact pin, so that 

 in the act of striking the coherer the hammer closes a 

 local circuit, and so causes a black mark to appear on 

 the chemical paper. Successive distinct marks can be 

 obtained in 0'0i7 second in this way, which is considerably 

 more rapid, I believe, than a decoherer was given credit 

 for. 



There is not sufficient time to show an actual trans- 

 mission by wireless, and I should like to make it clear that 

 only sketches of the simplest character are at present 

 I being transmitted ; but, as you will see from the result 

 thrown on the screen — a simple portrait of 

 His Majesty the King — the images are 

 recognisable, and merely require slightly 

 more detail to make them quite comparable 

 with the early results in line obtained by 

 Prof. Korn's telautograph. 



Another result shows a plan transmitted 

 by wireless ; here an island is seen repre- 

 sented, and a lighthouse — or it might be 

 a fort — and by means of letters the posi- 

 tions of sections of an army on the island 

 are supposed to be designated, while the 

 shaded portion might mean that the 

 " enemy " is in that part of the island. 

 Such plans as these could be drawn direct 

 in shellac ink on a slip of metallic foil, 

 placed upon a portable machine coupled 

 to a portable military wireless set, and 

 communicated from one section of an army 

 to another. The small portable machines 

 I have already shown are used for the 

 wireless transmissions, and they possess 

 the advantage that " tapping" of the com- 

 munications would be quite impossible. 

 It is for this reason that I think the 

 method would be of such value for military 

 and naval purposes ; even supposing that 

 anyone wishing to intercept a plan or written message 

 were to have an exactly similar instrument, with the same 

 dimensions, screw-threads, and so on, by merely altering 

 the rate of running by 5 or 10 per cent., according to 

 prearranged signals, the picture as received by the inter- 

 cepting party would be quite unintelligible and confused. 



We have already seen that in the telegraphy of a 

 picture by any system, accurate synchronising of the 

 sending and receiving apparatus is essential. Where a 

 metallic circuit links the transmitting and receiving instru- 

 ments together, the matter is an easy one, and we have 

 seen in what way it is effected. But when dealing with 

 wireless work, the question of synchronism becomes more 



T 



serious. I have employed two methods, each of which 

 appears to answer satisfactorily, and as they are very 

 important I will devote a few moments to their description. 

 The first method secures accurate synchronism independ- 

 ently of any wireless communication. You have already 

 seen how, in the ordinary telegraphic work, the receiving 

 cylinder is driven rather faster than the sending one, and 

 when it finishes up a complete turn too soon it is arrested 

 until the sending cylinder has caught it up, when the latter 

 sends a reverse current, which is responsible for its release. 

 But in the wireless apparatus both sending and receiving 

 cylinders are driven too fast, so to speak— that is, they 



