440 MEMOIRS OF THE AMERICAN ACADEMY. 



and the error and rate of the clocks or chronometer with the highest degree of 



I 



accuracy. 

 In the Coast Survey operations for obtaining differences of longitude between 



various points in the United States, by the Air Lines of telegraph, it is possible to 



make the longitude-signal record itself on the chronograph-sheets which register the 



time, at the station from which the signal goes and also at the station where it is 



received. But the electrical currents which are practicable with the Cable Lines are 



too weak to make this record at the latter station. They have barely strength to 



deflect a delicately suspended magnet, with an attached mirror, from which a beam of 



light is reflected upon a scale in a dark room. When the signal arrives and this 



deflection is seen, the observer touches his key and records the moment of its arrival, 



with as little loss of time as possible, upon the chronograph-sheet. 



The longitude signals arranged beforehand between Mr. Dean and Mr. Goodfellow 

 are of four kinds. 



I. At the beginning of a minute or any five-seconds mark on a mean-time chronom- 

 eter or watch, the observer at the first station sends a positive current, of exactly 



I 



wo seconds' duration, into the cable ; at the beginning of the next five-seconds mark 



* 



he sends a negative current of the same duration, to be followed at the third five- 

 seconds mark by a positive current of equal duration. After an interval of ten 

 seconds succeed four alternating currents in like manner, the first of which is nega- 

 tive. After another interval of ten seconds, three other alternating currents follow, 

 the first of which is negative. These ten signals make what is called a set, which may 

 be written thus: 



(P 2 )A(iV^)A(P 2 )_JL_(iV r2 )A(P 2 )A(iV^)A(P 2 )_i£_(iV^)A(P 2 )A(Ar 2 ). 



The object in dividing the set in this way into three groups is for the convenience of 

 identifying the individual signals with facility, when they are recorded upon the chro- 

 nograph-sheets. After an interval of ten seconds a second set is sent from the same sta- 

 tion. In recording these signals at the second station the observer, with his finger upon 

 his chronograph-key, watches closely the bright band, reflected by the mirror on the gal- 

 vanometer magnet, as it moves towards the end of the scale, and at the instant this band 

 darts towards the centre, he taps his key, and records upon the chronograph sheet the 

 moment when the cable begins to discharge its positive or negative current. After 



this, two sets of 



gnals are sent in the same way from the second station, and 



observed and recorded in a similar manner at the first station. The whole operation 



