294 HYDROGRAPHICAL SURVEYING [chap. xiii. 



by astronomical observations, we can find out by an exchange 



of signals what is the difference of time. 

 Chrono- In clu'onometric difference of longitude we have literally to 

 "l^"*^ carry the time from one place to the other. We ascertain the 



Difference -^ ^ 



of Longi- time at one place on a certam day, or, what is the same thing, 



tude. ^^.g £jj^ ^^^^^ ^Y\e Error of our clironometer on local time. 



Supposing for the moment that our chronometer is keeping 

 exact mean time, by carrying it to the other place and finding 

 out its Error on local time there, we can deduce the difference 

 of longitude by the difference of the two Errors. Thus, if our 

 chronometer is four hoiu's slow on mean time at A, and ■\\e 

 find when we get to B that it is three hours slow, we know 

 the difference of longitude is one hour, and that A is east of B. 

 Unfortunately, chronometers do not keep mean time, and 

 the problem is complicated by having to ascertain what time 

 they do keep, or, in other words, M^hat they gain or lose in 

 each day, which is called the rate. If we can find this, we 

 shall be able to get the difference of longitude just as accu- 

 rately as if the chronometer was keeping mean time, as we can 

 correct for this rate ; but here, again, chronometers are not, 

 and probably never will be, perfect instruments, and are 

 liable to change of rate, and it by no means stands to reason 

 that because a chronometer gains five seconds a day one week, 

 it will do so the next, especially when the sliip has been at 

 anchor during one period, and under way for the other, and 

 the temperature has not been invariable. 



Chronometric runs are therefore liable to the errors arising 

 from change of rate. To overcome this as far as possible, a 

 number of chronometers are carried instead of one, and, if 

 possible, what is called the travelling rate is obtained. 



Travelling If the rate of chronometers is obtained at a station A, and 

 Ave then go to another station, B, and obtain rate again there, 

 and apply the mean of these rates as the assumed rate of the 

 chronometers while being carried from A to B, we have no 

 guarantee whatever that this assumption is correct, as the 

 time employed in carrying the chronometers does not enter 

 into the calculation at all, and they may have been going 

 quite differently when the ship was at sea, mth the vibration 

 of the engines, motion of the ship, etc., to influence them, 

 to what they were when the ship was at anchor, besides the 



