1839.] Terrestrial Magnetic Force in Southern India. 231 



examine it, when the want of symmetry in one of its ends, fully justifies 

 the accuracy of the + 0°,37' found above. The other needles are, how- 

 ever, singularly perfect; so that the mean values here found belong to 

 needles Nos. 1, 2, and 3; and for No. 4 as far as concerns the poles a and 

 b' and a' and h ; but for the positions e and a' the correction = — 0°.37' 



and we may conclude, for e and h' = +1.1 



and for the remaining positions e' and h and e' and a = + 0.49 



In the actual determination of the dip, however, it is necessary to in- 

 vert the poles : so that putting p and ^ fur the corrections which apply 

 to needles Nos. 1, 2, and 3; we may 



in one position of the poles of the needle — read otF 



and in the inverted position of the poles we might read off 

 A + B 



where ^ = h the true dip, or, taking all the possible combinations 



of these six readings, the values for 5, resulting from either of the 

 needles 1, 2, or 3, and for several cases of No. 4, 



are 











A + B 



2 





or 



h ± 0. 0. 0 



A + B + 

 2 



P 



4 



or 



r5 ± 0.24.30 



A + B _^ 

 2 



P 

 2 



or 



a ^ 0.49. 0 



A 4 B ^ 

 2 



3^ 

 4 



or 



a r 1.13.30 



A 4- B ^ 



9, - 



£_ 



or 



a f 1.38. 0:— 



and since, in the remaining cases of needle No. 4, there are several 

 values entering, which the observations afford no clue to identify, if they 

 are — poles at e and a' or at e and 6', it will in cases of doubt be as well to 

 reject the observations of needle No. 4. With this view of the case, and 

 recollecting that among several independant observations, — in the greater 

 number of cases no correction whatever will be necessary — the poles 

 will have identified themselves with the extreme ends of the needle, — 

 we will now consult Table I, and endeavour to discover what propel' 

 compensations may be applied to the values there given, to obtain the 

 true values of b 



