.344 



HYDROGRAPHICAL SURVEYING [chap. xiv. 



Mean 



Harbour 



Rates. 



Adapta- If we obtain Error before leaving A, and after some days 

 Fore-°^ call at B, whose difference of longitude from A is known, and 

 going. there obtain Error again, Ave get a very good sea rate for the 

 subsequent part of our voyage, wliich we can utilise to deter- 

 mine the position of C, any third place at which we may here- 

 after soon call, with a probable better result than by means of 

 harbour rates. If absolute altitudes are used, they must, of 

 course, be both either a.m. or p.m. 



This method is especially useful for navigational purposes. 

 Suppose a ship to leave Portsmouth and to call at Gibraltar for 

 a few hours only. Error can be obtained, and by means of the 

 knoAATi difference of longitude a sea rate deduced, which will give 

 a better landfall for Malta than the harbour rates at Portsmouth. 

 When our voyage is simply from one port to another, and 

 we wish to find the meridian distance between them, we must 

 depend mainly upon the harbour rates ascertained before 

 departure and on arrival. 



The ordinary and rougher method is to assume that the 

 rate has changed uniformly from the rate of departure to that 

 of arrival, and that therefore the mean of the two rates will 

 represent the mean rate during the passage. We believe that 

 (owing to the many causes of variation impossible to formulate) 

 in most cases, and especially where temperature has been, in 

 the clironometer room, fairly uniform, this method will give 

 as good a result as any other ; but where temperature has 

 changed much, the result of long meridian distances with 

 such rates will have but very little value, and that a correction 

 for temperature will much improve the result, if we can apply it. 

 French naval officers have done much in working out this 

 Cj[uestion, and Captain Shad^ell gives their separate theories 

 and formulae. To our mind the method of M. Mouchez is the 

 most practical ; and not undertaking to enter into the question 

 of acceleration, nor depending on observatioixs on the watches 

 while in the observatory, it is more adapted to actual work. 

 Mouchez s Mouchez proceeds on the assumption, which is near enough 

 to truth for the method, that the rate varies uniformly with. 

 the temperature ; but in working on this hypothesis, Ave must 

 not forget that for each clironometer there is a point of tempera- 

 ture at Avhich the rate h at a maximum, and that the sign of 

 the variation Avill change as Ave pass it. 



Rule. 



