'^IS Olsorcaiions on the Direction and Intcnsily of the [Ariat 



anil thaf, to trat-e the law of its vaiiation nmlor tlicso ciivunistaiico'^, 

 obstM- vat ions of far nioro tlian ovdiiiary accuracy should be otuploycd. I 

 liati how over other motives in view in uuikiug these observations tban 

 that of diseovering from them the law of magnetic intensity. My object 

 %vas, not to discover the hiw, but — assuming the hiw according to theory — 

 how far it was transgressed by incidental or local causes : for,with regard 

 to the latter, our track along the Coromandel coast, surrounded by ex- 

 tensive sandy plains, would form a fair field for exhibiting the elfect of 

 local disturbances, when contrasted with the vast granitic formations 

 and elevations to be met with in and near to Travancoro and along the 

 Western coast generally. 



With thus much by way of explanation, I will now lay before the 

 reader the whole of the observations, even those which I shall even- 

 tually propose to reject, in order, that no more, or less importaiice may 

 be attached to them, thati they are fairly entitled to. I will here mention 

 one trifling circumstance with regard to Mr. Caldecott's needles; namely, 

 that, with a view to try if their intensities might not be rendered at once 

 invariable, before leaving Trivandrum I submitted each of them to rather 

 rough usage by hard friction. The effect of this was, that a diminution 

 of intensity ensued at once — to the amount of about 20 seconds in per- 

 forming 100 vibrations. How far this might be carried, is a matter of 

 interest for which I may not pause here, save tu remark, that, as far as 

 my experience goes, a needle may by this means at once be brought to 

 a fixed intensity; of which more presently. It only remains for me to 

 add, that the same observers and the same stations as those mentioned 

 in the dip observations, apply to these. 



Corrections to he applied to the Intensify Observations. 



The corrections to be applied to the observed time of one of these 

 needles, performing 100 vibrations, are three in number. 



1st. — On account of R the rale of the chronometer ; which renders ne- 

 cessary the correction 



, 86400 \ 



rate = I-(I X gg^^^Q ^ r J 



2d. — On account of arc. To reduce the time of vibration performed in 

 a circular arc commencing from A°, and terminating at to that which 

 would have been noted had the vibrations been performed in an infiiiitely 

 small circular, or a cycloidal arc, the correction is 



sin. (A -f «). sin. (A — «) 



corr. for arc = + I X oo a/t n '• \ 1 T 



^ 32 M (log. sm. A — log. sin. «) 



where M — the logarithmic modulus = 2,302585 ; and A and a, througli- 



out — have always been 20° and respectively ; or correction for arc 



= 4- (I X,0022). 



