REPORT ON 



I I/ 1 /^0 + ^10 ^8 + ^24\ 



A 2 ~2 / 



1/^1 + ^17 ^"f~'Vfl 1/^5 + ^21 ^13 + ^29\ 



f I (, 2 2~ / "" * V 2 2 / 2 



l/^2 + 5l8 ^10 + ^26\Q, 1/^6 + ^22 ^ 14 + ^ 30\ o 



"*\ 2T" "2 / 4 ~ 2 2 / 4 



1/^3 + ^19 ^ll + ^27\o 1/^7 + ^23 ^IS + ^Sl 



+ * \ 2~ ~2 / ' "" 2 \ 2~ ~2^ / 



But the deviations about to be discussed were all observed, not on the compass 

 points, but on the correct magnetic points. Treating them in the manner which 

 has just been described, we obtain the approximate coefficients A^ B^, C } , D^ E^ 

 which belong to the correct magnetic points. Then, from equation (11) we get, 

 going to terms of the third order inclusive, 



5 = 21 (14) 



+ (33 + 31 G) sin f + (G 21 S3 cos 



- 



_S(V_ 

 .^ a _|_(23 a -S 2 )Dj 8in4^+ - 



cos 



where 5 is expressed in terms of the arc which is equal to radius. If we suppose 

 the complete expression for 5 to be 



5 = ^1 1 + J5 1 sin^+C'! cos^ + D, sin 2^ + ^, cos 2 (15) 



+ F, sin 3 + G l cos 3f + H, sin 4 + 7^ cos 4f 



sin 6 



