488 PROFESSOR CHALLIS, ON THE DETERMINATION OF THE LONGITUDE 



action takes time to pass along the wire, the effect of the retardation on the determination of 

 longitude is eliminated by transmitting the signals in nearly equal numbers from the two 

 positions, and taking the mean of the results given by the two sets. The following is 

 Mr Airy's " Scheme for the making and observing of Galvanic Signals for the Comparison 

 of Distant Clocks," drawn out subsequently to the trial of the transmission of signals 

 between Greenwich and Dover. 



1. Suppose an hour to be granted for the use of the telegraph wire. The hour is 

 divided into four quarters, and during the first and third quarters signals are sent from 

 Greenwich, and during the second and fourth quarters signals are received at Greenwich. 



2. Consequently the fittings at both ends of the line must be generally similar. There 

 must be a battery (one of 72 pairs of plates was very vigorous from Greenwich to Dover), 

 and means of putting the battery out of circuit for receiving signals, and of putting it in 

 circuit by starts for giving signals. The mechanism for this purpose of the common needle 

 telegraph answers perfectly well. There must also be a signal-needle in another room, or 

 in a distant corner of the same room, at which the time-observations of signals going both 

 ways are made. The separation of the signal-giving place from the time-observing place 

 has been found to be indispensable for accuracy. 



3. The signal is rather startling and requires an observer in a good state, of nerve. 

 Considering that a half-second beat comes so quick as to endanger the requisite coolness, 

 it would be best, if the means are at hand, to erect a clock near to the signal-needle. 

 But if not, a chronometer may be used : at Greenwich it was done every time with fair 

 success. 



4. The signals are made in the following way. The signal-giver fixes in his mind 

 on an arbitrary number of signals to be given in the next batch : not fewer than three 

 and not more than nine. Suppose the number to be five. Then he gives five warning 

 signals at intervals of about ]-| or 2 seconds (to prepare the mind of the signal-receiver), and 

 follows them with five time-signals at arbitrary intervals, not less than 10* nor greater than 



15 s . Graphically a batch is represented Then he waits 20 or 30 



seconds before making another batch. 



5. At each end there must be three persons. The signal-giver A will have nothing 

 to do but to give signals, as above described, for half the time. The signal-watcher B will 

 take seconds from the clock, count beats and observe the start of the needle, and enter the 

 times in his book during the whole time. The assistant C (who is to be a very quiet person), 

 must look over the shoulder of B, and count the warning-signals and say "five signals 

 coming." He must quietly count them as they come, and after the entry of the fifth must 

 say " batch finished." This assistance is a great relief to B. 



6. In the successive quarters of hour, the needle has been thrown right in the first and 

 second, and left in the third and fourth. 



