1840.] Fort William, and Futtehghur. 85 



As Troughton's directions for using his circle are not uni- 

 versally known, I will here extract from them a few lines, in 

 which he plainly states its advantages, when compared with 

 the sextant ; they are chiefly these : — 



(i The observations for finding the index error, are ren- 

 u dered useless ; all knowledge of that, being put out of the 

 " question, by observations both forwards and backwards. By 

 u the same means the errors of the dark glasses are also cor- 

 u rected, for if they increase the angle one way, they must 

 " diminish it the other, by the same quantity. This also 

 " perfectly corrects the error of the horizon glass, and those 

 ie of the index glass, veiy nearly. But what is of still more 

 " importance, the error of the centre is perfectly connected, 

 " by reading the three branches of the index, while this 

 u property, combined with that of observing both ways, proba- 

 " bly reduces the errors of dividing, to one-sixth part of 

 " their simple value. Moreover, angles can be measured as 

 u far as one hundred and fifty degrees, consequently the sun's 

 " double altitude may be observed, when his distance from 

 " the zenith is not less than fifteen degrees, at which alti- 

 " tude the head of the observer begins to intercept the rays of 

 " light, incident on the artificial horizon, and of course if a 

 " greater angle could be measured, it would be of no use 

 " in this respect." 



Mr. Troughton has not noticed a farther great advantage, 

 in there being no need to take the index error of the circle, 

 as there is with the sextant; the finding this error with the 

 latter, as it is generally done by measuring the sun's diameter, 

 on each side of the zero, is well known in these hot coun- 

 tries to be a most painful, as well as a tedious and uncer- 

 tain operation, and we measure only on a small part of the 

 arc the glaring disk of the sun, through the stained glasses, 

 which we see under a very different degree of brightness, 

 from that under which we take the contact of the moon 

 and sun or stars, and this index error ought, with the very 

 best sextant, to be rigorously examined at each observation. 



With the circle the correction for the zero point is included 

 in the observed distances on both arcs, and given on six parts of 



