54 THE TRANSATLANTIC LONGITUDE. 



VII. 



LONGITUDE SIGNALS BETWEEN FOILHOMMERUM AND HEART'S CONTENT. 



* The method of giving and receiving the signals for longitude, betvi^een Foilhom- 

 merum and Heart's Content, was that prescribed in the programme which I had 

 prepared before leaving home. Three series, of two sets each, were exchanged on 

 every occasion ; each set consisting of ten signals alternately positive and negative, 

 at intervals of about five seconds, except that the fifth and eighth were preceded 

 by pauses of ten seconds, which was also the interval between the two sets. The 

 purpose of this arrangement was to discover whether the velocity of transmission 

 was perceptibly afi'ected by a longer thne being allowed for the cable to recover its 

 electrical equilibrium, and also to facilitate the identification of the individual 

 signals. Some slight convenience in the practical details also arose from the cir- 

 cumstance that each set occupied one minute, and that each series consisted of ten 

 positive and ten negative signals. Those signals were considered positive by which 

 the platinum was put in connection with the cable and the zincode with the ground. 



In receiving the signals, the observer (Mr. Dean at Newfoundland, and myself 

 at Valencia) watched the deflections of the light-spot, while his thumb rested on 

 the button of a delicately adjusted break-circuit key, which was pressed at the 

 instant in which the deflection was perceived. This instant was thus recorded 

 upon the chronograph, after a certain amount of delay, which we will call the per- 

 sonal error of noting, and which depended upon a considerable number of influences 

 to be discussed hereafter. The keys by which the signals were transmitted were 

 made by the American Telegraph Company, imder the supervision of Mr. Dean, 

 and are constructed according to the arrangement devised by Prof. Thomson for 

 the Atlantic Telegraph, in such a manner that pressure upon one button produces 

 a positive, and upon the other a negative signal, while no current flows at other 

 times. To this arrangement an additional contrivance was applied by which the 

 local circuit to the chronograph passed through the same key, and was interrupted 

 by pressure upon either button, so that every signal transmitted through the cable 

 was recorded upon the chronograph at the station whence it was sent. 



It is thus manifest that the times of sending the signals were accurately recorded, 

 while the times of receiving signals were recorded after an interval of time depend- 

 ent on the personal error of noting, and inseparable from the time of transmission 

 through the cable, except by some independent means of measurement. If this 

 interval were the same for both observers, it would be eliminated entirely from the 

 longitude and merged with the time of transmission. Otherwise it would affect the 

 resultant longitude by one-half the difference between the personal errors of noting 

 for the two observers. Happily it proved to be very nearly the same for Mr. Dean 

 and myself, and also measurable ; so that it has been possible to eliminate its inffu- 

 ence from the measure of velocity, as well as from the longitude. 



If now we denote the clock-times at Valencia and Newfoundland by T and T' 

 respectively, the corrections for reducing these to the true sidereal time by At and 

 At' , the time required for transmission of the galvanic signals by x, and the longi- 



