88 Differences of Meridian at Madras, [No. 97- 



In determining latitudes, the reflecting circle is most useful 

 to the geographical surveyor and navigator. By no instru- 

 ment can so many good observations be taken in so short 

 a time, the meridian altitudes of the sun and stars, one day 

 or night, taken by the readings of the three branches of one 

 arc are corrected by those of the next; but by a still more 

 rapid, and equally accurate process in one day, a sufficient 

 number of circummeridional altitudes of the sun can be taken, 

 and reduced to the meridian. During ten minutes on each side 

 of noon, ten or twelve double altitudes may be well taken 

 from the artificial horizon, marking the horary angles by the chro- 

 nometer, and at night, many stars may in like manner be 

 calmly and well observed. They are best selected on both sides 

 of the zenith, and the time from noon may be extended in 

 proportion to the slowness of the star's motion ; with the 

 pole star to a great extent at sea, but on land, in geodesic 

 operations, it may be extended to half an hour on each side of 

 the meridian. Of course, as in a lunar distance, the observations 

 must be. taken on the right and left arcs alternately, or on 

 equal numbers before and after the meridian passage. A stand 

 is sometimes used with the circle, but I always found I could 

 work quicker and better without. In oblique distances a support 

 for the elbow is desirable, but in taking altitudes, the best 

 way is to sit on the ground, the back being supported by a 

 Hindostanee morah, or some such thing ; this posture gives 

 to the hands perfect command over the instrument ; also re- 

 member that when the glass roof is used over the mer- 

 cury it should be reversed at each contact. Circumstances 

 may prevent observations with the circle being obtained on 

 both the arcs, " in such cases the instrument may be used 

 in the manner of a sextant, and the index error applied, 

 with this advantage over the sextant, that the index error, 

 as well as the observed angle which it is to correct, is read 

 on three verniers instead of on one, as with the sextant ; also 

 the index error may be taken on two small stars, or other 

 well defined objects, subtending a greater angle than the sun's 

 diameter, and the usual painful operation of measuring it 

 avoided altogether. With the circle also, which is called at sea 



