THE TRANSATLANTIC LONGITUDE. H 



circumpolars should be reversed upon ; and stars as far north as 80' should be observed by the old 

 method of eye and ear, instead of the chronograph. 



Whenever possible, sets of observations should be made at least twice during the night, each set 

 consisting of not less than three circumpolars (not all at the same culmination), and three tinie-stai# 

 north of the equator, together with any southern time-star which may be convenient. A set of 

 observations should always precede, and another set follow, the exchange of signals, when the 

 weather permits. 



One or more of these sets should be computed promptly, that observers may constantly be 

 • acquainted with the condition of their instruments. The azimuth error should never remain for 

 more than a day larger than 0'.2, nor the collimation error larger than OM. For the field computa- 

 tions it will suffice to read off a single tally for each star. 



The amount of battery-power and condition of the wire is always to be noted when telegraphic sig- 

 nals are exchanged ; also any indications of aurora. 



Hearfs Content-^ Calais. 



So soon as the instruments are in adjustment, the exchange of clock-signals should commence, 

 and it should be continued nightly, whatever the weather, until the operations for trans-Atlantic 

 longitude are completed at Heart's Content. At the time of the exchange, Calais should notify 

 Heart's Content whether it can determine the clock-correction on the same night; and should 

 transmit the correction deduced for the time of the signals sent on the preceding night. 



To exchange clock-signals, put the Calais clock into circuit two or three times, for not more than 

 half a minute at each time and at intervals of af^least a minute, while the Heart's Content clock is 

 graduating the chronograph. Arrange the time for putting on the Calais clock, so that the record 

 of 0' shall be included in the series of its signals. It is very desirable that both chronographs 

 should record this comparison, but if this should be found impossible, the Heart's Content chrono- 

 graph is the proper one to keep the record. If any confusion is likely to arise as to the precise 

 seconds recorded by the Calais clock, this can be readily obviated by making a couple of quick taps 

 immediately after 15', 30', or 45' of the clock-time, entering this fact upon the day-book, and com- 

 municating it to Heart's Content. 



Valencia — Hearfs Content. 



1. For this determination, three nights' exchanges through each cable will suffice, provided the 

 clock-corrections are well determined at each station, before and after the exchange. Should cir- 

 cumstances be especially favorable on any occasion, there is no reason why work should not be 

 done with both cables on the same night, thus reducing the requisite number of nights to five. 



2. The times at which exchanges will be made must necessarily depend upon the convenience of 

 the Telegraph Company; but the hours between 10 P. M. and 6 A. M. are preferable. (All civil 

 times in this programme are understood to be Greenwich mean times.) Whenever exchanges are 

 to be undertaken, Valencia will notify Heart's Content as early as 6 P.M., if practicable, naming 

 the hour when this can be done. Should no such notice be received by midnight. Heart's Content 

 need not feel obliged to attend farther. 



3. At the appointed hour, Valencia will telegraph the word Gould, as a notice that all is ready; 

 and upon the reception of the word Dean in reply, will begin the signals. 



4. The exchange of signals will be effected as follows : — 



a). Beginning with a positive current, sets of alternate positive and negative signals will be made, 

 each signal consisting of a single tap half a second in length. The first group will consist of four 

 taps, at intervals of five seconds. Then, after a pause of ten seconds, will follow a group of three 

 taps, five seconds apart ; and, after a second pause of ten seconds, yet another group of three taps 

 at five-second intervals ; these ten taps, in three groups, constituting a " set." The arrangement 

 of the set will then be thus : — 



P=.N,,P,.N„, P,,N33Pi.. N,,P,,N 

 and each set will occupy one minute. 



6). Two such sets, following one another at an interval of ten seconds, will be sent fir.st from 



