Mr Sang on Optimum Surveying. 343 



may occur among equation 3 as well as among equation 2. If A 

 become in turn an observed station this will happen, in which 

 case the two equations involving d . Ion A must be added to- 

 gether. 



4!th, Depending on latitude of station observed from ; d lat A. 



AN f a* 132 



2 . __ J ^.cos lat N . cos lat A , cos (Ion A— Ion N)+ — sin lat N . sin lat A 



_ (''-^Tt(f2 1atAy^_ «^^ ,„, 2 lat A I = O 

 (A)3 (A) J 



And, 5th, Depending on latitude of signal ; 5 . lat N. 

 2. /f^li^^=l5!)sinlatN.coslatN2— -^ sinktNJ X 



X ^ sin lat A . cos (Ion A — lonN) — 1- sin (Ion A — Ion N) — ^— i- > 



+ 2 . 1 ^^^=^^ sin lat N^ cos lat N --^-^cos lat N I cos lat A . ^i^^ 



= 0. 

 To equations 4 and 5 a remark similar to that made con- 

 cerning 2 and 3 applies. 



The observations of the bearings are not, however, the only 

 data of a survey: the observed longitudes and latitudes of 

 various stations have to be combined with these. Using capital 

 letters to indicate the observed latitudes and longitudes, we 



have 



lat A — Lat A 



for the error in latitude at the station A; and for the error 

 which this will cause in the position of A, we must add 

 as factor the radius of curvature of the meridian: so that, 

 roughly estimating this radius, and combining it with the pro- 

 bable error caused by inaccuracies in the zenith sector, we ob- 

 tain a probability E of inaccuracy. In this way 



(latA— Lat A\^ 

 E / 



is the measure of inexactitude, and thus 



latA — Lat A 



^ E^ 



falls to be added to the equation depending on d . lat A. Or, 



