272 



HYDROGRAPHICAL SURVEYING [chap, xn 



Latitude 

 by Obser- 

 vations 

 " Abso- 

 lute." 



But more 

 Difficult. 



Elimina- 

 tion of 

 Errors in 

 Observa- 

 tions for 

 Latitude. 



LATITUDE BY CIRCUM-MKRIDIAN ALTITUDES OF 



STARS. 



Determinations of latitude are more simple in one respect 

 than those for longitude, as they are " absolute " — that is to 

 say, they depend solely upon themselves ; whereas longitude 

 has to be obtained by the difference of two sets of observa- 

 tions at two different places, and is further complicated by 

 the eccentricities of the clironometers upon which, when there 

 is no telegraph, we have to rely. 



But, on the other hand, the observations required for correct 

 latitude are more difficult to take, as, to arrive at anything 

 like an exact result, we must use stars, and each step of the 

 observation of these in an artificial horizon is rendered less 

 easy by the fact of their being made at night. It is much 

 easier to become a good day observer than a good night one. 



The errors to be eliminated as far as possible in observing 

 for latitude are, firstly, errors of observation ; secondly, instru- 

 mental errors, as centring error, index error of sextant, error 

 caused by refraction in the rays passing through the glasses of 

 the roof of the horizon, etc. ; tliirdly, atmospheric refraction, 

 which varies much, and for which no known rule of correction 

 thoroughly suffices ; foiu-thly, personal errors, caused by each 

 individual's mode of observing the contacts. 



Another source of error in observations both for time and 

 latitude, when using sextant and artificial horizon, is due to 

 there being no means of ensuring that each observation is made 

 precisely in the same spot in the artificial horizon. Great care 

 is therefore necessary in this respect. 



Roof error is minimised at altitudes between 40° and 50°. 



Errors of observation are eliminated by taking as many 

 observations of altitude as we can, and we must therefore 

 observe off the meridian, or what are known as circum-meridian 

 altitudes, which consist in observing from a short time before 

 the meridian passage to a short time after it, and adding a 

 certain correction to each altitude to make it equal to the 

 meridian altitude, and thus get a mean meridian altitude, 

 which, if we can calculate the correction exactly, will be of 

 much more value than actual observation on the meridian 

 only. 



