>rAfA'ETIC SURVEY OF J.VPAX. 1^'7 



1\('1(1 for tlic Menu Stati(in. Tlie lengtlis in kil<tinctrc.s (»f erne niimiic 

 ill :ii-c oi' Infitudê and lonoitnde at that station are : — 



1 minute of Latitude 1.8.5 kilometres. 



1 ,, ,, Longitude L49 „ 



This transformation of units l)eing effected, it was llien easy to 

 fuid tlie angle a1 which any given iso-magnetie line at the ]\[eaii Sta- 

 tion cuts the meridian line, and the greatest rate of change per kih^- 

 metre of the particidar element at that Station. 



L — The Dip. 



The fifty selected observations, when combined hy the metliod 

 of least squares, ga\e the following formula expressing the Dip ( (?) 

 in terms of the co-ordinates : — 



e = 50°28'.6 + (1.141 9 - .1556 ;,)' 



9 and Â are tlie lafiru(k' and longitude co-ordinates referred to 

 Ihe mean station (36° 30' X. Lat.. 137^ 0' E. Long.) and measured 

 in minutes of arc. 



In the annexed table, Table XL, the values of the Dip as observed 

 at all the stations are s'iven, and alons; side of thom the values as 

 calculated from the Formula. 



The selected stations, on which the calculation depended, are 

 indicated by having their numbers printed in heavier type. includ- 

 ing the other stations serses to indicate to what degree of approxim- 

 ation the f)rmula a])plies to them also. This will be best shown hy 

 giving the probable cn-or of a single observation in the different cases. 



If we take all the -Sf stations, the j^robnlile error comes out 



± J 1 .OO. 



If we neglect Xos. .'^)!), 11. and 78. whose differences ai"(^ verv 

 larg«' indeed, the ])rol)able error hecomes ± S'.5I. Xos. ,')i) and 41 



