1844.] 



in southern India. 



197 



Under these circumstances I think we may safely conclude 

 that more reliance is to be placed in the smaller result of 2'6" 

 than in the large one of 6' 6". 



That the observations now made are not materially in error 

 is proved by the accordance of the results taken with differ- 

 ent circles and Needles. 



Thus on the 19th April 1844 the Dip taken with a circle 

 in the possession of Mr. Taylor, (the same used in the obser- 

 vations made in 1843 and 184-1) was found to be 7° 25' 41" 

 This result being the mean of 4 different Needles. On the 

 same day the Dip found by a different Instrument and Nee- 

 dle ■ (The one always used in the Stationary series) was 

 7° 26' 00" or differing from the other by only a few seconds. 



The Dip taken with the last Instrument and Needle on 

 the 16th July 1844 was found to be 1° 21' 22". On the 

 following day with another circle belonging to the Magnetic 

 Observatory the mean of 3 needles made it 7° 28' 33", dif- 

 fering from the other by a little more than 1' 



There is one circumstance arising from a comparison of 

 these two sets of observations which appears worthy of at- 

 tention. Tracing out the line of the' magnetic Equator first 

 from the observations made in 1838, and then from those 

 in 1833 34, we obtain the figures shewn in the accompany- 

 ing Diagram. 



Were we only to regard one of these we might suppose 

 that the devious course of this line was simply owing to 

 errors of observation, but when we come to compare the two 

 we cannot help being struck by their extraordinary similari- 

 ty which it would be unreasonable to attribute to the effect 

 of accident. The only way by which I can account for it 

 is by supposing that the observations are sensibly effected 

 by local attraction. The observation taken at the Eed Hills 



