TELEGRAPHY OF PHOTOGRAPHS BAKER. 



259 



and shade of the picture is made up of masses of thinner or thicker 

 lines, with clear spaces in between. 



As the stylus traces over such a photograph, its contact with the 

 metal base is interrupted every time one of these fish-glue lines comes 

 beneath it, and for such a time as depends, of course, on the width of 

 the line. The transmitting instrument thus sends into the telegraph 

 lines a series of electric currents whose periods of duration are deter- 

 mined by the width of the lines composing the photograph. 



A similar stylus, S 2 , traces an exactly similar path over a revolving 

 drum in the receiving instrument, but round this drum is wrapped a 

 piece of absorbent paper impregnated with a colorless solution, which 

 turns black or brown when decomposed by an electric current. 



What happens then is that every brief current which passes 

 through the paper causes a mark to appear on it. The width of the 

 mark depends on the duration of the current — or should so do — so that 



Fig. 1. 



you will see that these marks gradually combine to recompose the 

 photographic image. 



This method is all very well in the laboratory, but when we come 

 to try it over a long distance the capacity of the line at once causes 

 serious interference. It is well known that if a current be sent to 

 some apparatus, such as a telegraph, from a distance, the current 

 having to pass through long wires the capacity of which is appre- 

 ciable, a certain time is taken for the current to charge the line, and 

 the line discharges itself into the apparatus with comparative slow- 

 ness. If the circuit be closed by means of a Morse key, the time of 

 contact at the key being a sixth of a second — a common time of dura- 

 tion of a short tap — the discharge of current from the cable would be 

 considerably longer than one-sixth of a second. AVhen, therefore, 

 we are sending signals through the line at the rate of 175 per second, 

 it is not difficult to see that every signal will run into the next dozen 



