348 ABSTRACTS OF SCIENTIFIC PAPERS 



sending dots so rapidly that no variation in the received current 

 could be detected, notwithstanding the sensibility of the receiv- 

 ing instrument. When dots and dashes were sent alternately, 

 and sufficiently slowly to be separately observed, the end of the dot 

 curve was found to go lower than the end of the dash curve, a 

 fact which limited very much the speed of sending possible with 

 an instrument of the Morse type. This was when the interval of 

 connection to earth was the same after both dot and dash, and it 

 was pointed out that this defect in the signals could be remedied by 

 making the interval of earth contact after the dash longer than that 

 after the dot, in a definite ratio. Further, to prevent the received 

 current from sinking during the longer intervals required for spaces, 

 it was suggested that each of these should be produced by a number 

 of very rapid short applications of the battery, with short earth 

 contacts between. To carry out these ideas automatic sending 

 became necessary, and Mr. Jenkin showed that the suggested system 

 was practicable by an experiment in which, in place of an ordinary 

 sending key, a strip of paper was used as sender, perforated with 

 long and short holes which allowed electric contacts to be made for 

 the desired lengths of time. By this means common Morse signals 

 were successfully transmitted through the cable with much greater 

 rapidity than would have been possible in the ordinary mode of 

 sending. [This ingenious system of automatic sending was developed 

 in much detail and formed the chief subject of a joint patent of Sir 

 W. Thomson and Mr. Jenkin in 1860. The system did not, how- 

 ever, come into use. It was rendered less necessary by the adoption 

 on all long submarine cables of Sir W. Thomson's mirror galvano- 

 meter, and subsequently of his Siphon Recorder, with the result that 

 any sensible fluctuation in the received current could be followed 

 and interpreted as a signal, irrespective of its amplitude and of its 

 position on the curve.] A mathematical investigation by Sir "W". 

 Thomson of the relation which the amplitude of fluctuation in the 

 received current bears to the period of the sent currents, when 

 these are sent at regular intervals with earth contacts between, is 

 given as an appendix to the paper. The experiments included 

 observations of the speeds of sending of uniformly spaced short 

 currents on various lengths of cable, at which the amplitude of the 

 signals became (1) so small as to escape detection and (2) just capable 

 of being received without confusion. The results agreed very exactly 

 with the deductions from Sir W. Thomson's theory. 



