272 W. E. AND F. B. COOKE. 



brevity, let us call the clock circuit c and the radio circuit 

 r 4 aod note that it is essential that the duration of the 

 periodic closing of c shall be greater than the duration of 

 the r signal. E.g. if the r signal occupies O'ls, then the c 

 must close every second for 0*15s or any period longer than 

 this. There need practically be no limitation to the length 

 of the r signal, provided that the c impulse is a trifle longer. 

 There will thus occasionally occur a condition when r tries 

 to come through but finds c closed and consequently no 

 path available, and we shall then cease to hear the r signals 

 for a time. Taking the above figures, this overlap will be 

 0*05s, and with r gaining about O'Ols per second the r 

 signals will find their path blocked for five consecutive 

 seconds, and we shall have silence for that period. In 

 addition, the few r signals immediately preceding this 

 silence will be cut short. They will commence to pass but 

 almost immediately their path will, he blocked by the clos- 

 ing of c and we shall notice a gradual diminution until they 

 finally cease. The coincidence may be taken either as the 

 last audible signal or a second later. Theoretically it never 

 occurs at any exact second, and we must adopt either the 

 one immediately preceding or following. Practically it 

 makes no difference which we adopt, provided we act 

 systematically; and then even the third place of decimals 

 will in the long run be just about correct. To those 

 accustomed to chronographs the following diagram may 

 help to make the matter clear. The whole sequence is 

 purposely exaggerated. 



r- i m n n n n 



_n n n_ 



3 



These two irregular lines show the records as they would 

 appear if one could transfer them to a chronograph. 



At 1 and 2 the signals come through without interruption. 

 At 3 the signal starts but is broken when c closes. At 4 



