SUBMARINE TELEGRAPHY 239 



and in copper, can only be about two and a half times as good a 

 speaking instrument as the 1858 Atlantic, on which only two and 

 a half words per minute had been obtained. 



In order to produce a succession of distinct and legible 

 signals, it is not necessary that the wave should reach its maxi- 

 mum and fall to its minimum at each signal. If the sendincr 



O 



battery contacts are changed or reversed before the full height 

 of the wave is reached the wave is not obliterated, it is simply 

 diminished; if battery contacts, alternately with one and the 

 other pole of the battery, succeed one another with considerable 

 rapidity, say three reversals every second, or ninety dot-signals 

 per minute, the waves will be reduced to say 10 per cent, of 

 their maximum ; but if we can render these little waves visible 

 they may be interpreted as legible signals. The old Morse 

 system, which simply indicated a blunt Yes or No, could not 

 show these little waves or follow them in any way. Sir 

 William Thomson's reflecting galvanometer does render these 

 little waves legible, even when they are no larger than 1 per 

 cent, of the maximum current. The received current deflects 

 a tiny magnet to and fro. A little mirror swinging with the 

 magnet reflects a spot of light on to a distant scale, where by its 

 oscillations the spot of light indicates movements of the magnet 

 too small to be directly seen. The little swinging magnet fol- 

 lows every change in the received current, and every wave, great 

 or small, produces a corresponding oscillation of the spot of light 

 on the scale. These little oscillations, produced in due order 

 are easily read by a practised clerk no one knew how easilv. 

 The sending arrangements were designed to produce perfect 

 regularity in these little waves, making them as sharply defined 

 as possible ; but just as by a practised eye handwriting can be 

 read in which there is no single clearly formed letter, so it has 

 been found that a clever clerk will instinctively, as it were dis- 

 entangle most irregular oscillations of this little spot of li^ht 

 into the letters and words they should represent, and from this 

 cause the greater part of the sending gear has been found un- 

 necessary, and yet the bold estimate of eight words per minute 

 has been exceeded. The addition of a single simple little in- 

 strument allows four times the number of words to be sent 

 through the Atlantic that could te sent through the Malta- 



