1834.] 



of the Madras Mural Circle. 



405 



bad been previously adjusted to subtend as nearly as possible an angle 

 of 90° with the former ; the difference betwen the reading now obtained, 

 and the first reading, gives the exact angle between the two, 4- or — 

 error of division. The telescope being now released from the circle was 

 again directed to the first-named telescope ; again clamped to the circle; 

 and, by means of the movable wire, a bisection again made ; releasing 

 the circle, it was with the attached telescope again turned till the 

 telescope bisected the wires of the smaller telescope, when 1 80° nearly, 

 or double the angle subtended by the pairs of cross wires, was read off; 

 pursuing this course after five readings we arrived at the point of depar- 

 ture : now the difference between the first and last divided by four, or 

 the true value of the angle, being added to the first reading gives the 

 second ; being again added gives the third, &c. these compared with 

 the actual reading give the error of division of the points 0°, 90°, 180 o , 

 &c. thus : — 



A. 



B. 



Mean of 

 A and B. 



Mean An- 

 gle. 



Interpolat- 

 ed. 



Error of 

 division. 



1 >> 



3 0.0 



1.1 



5.2 



12.2 



17.4 



20.9 



0.55 



5.20 



11.45 



16.40 



21.90 



21.90—0.55 



0.55 



5.89 



11.23 



16.57 



21.90 



0.00 



90 5.2 

 180 10.7 

 270 15.4 

 360 22.9 



4 

 // 



= +5.34 



—0.69 

 +0.22 

 —0.17 

 0.00 



From the mean of ten repetitions in this way the errors came out, 



r and 180 = 



0.00 



90 and 270 



—0.45 



180 and 360 



—0.03 



For the error of divisions intermediate between the above, I placed 

 Dollond's 5-feet Telescope as before, and by means of two wooden 

 supports, A. B., a 42-inch Achromatic, by Dollond, was rv-i p^~| 

 supported immediately above, so that a pair of cross lines L/vJ \_z\ 

 fixed in the focus, subtended an angle of 30° with the cross-lines 

 fixed in the other Telescope. I then made the following measures • 





A 



B 



Mean of 



Mean An- 



Interpolat- 



Error of 









A and B. 



gle. 



ed. 



division. 







• » 



u 



ii 



a * 



a 



// 







0.0 



1.9 



0.95 



18.50—0.95 



0.95 



00 



30 



5.0 



7.0 



6.00 



3 



6.80 



—0.80 



60 



12.0 



12.6 



12.30 





12.65 



—0.35 



90 



18.6 



17.5 



18.05 



=5". 85 



18.50 



—0.45 



In a si in 



ilar manner. 







// // 







90 



0.0 



59.2 



59.60 



1.88—0.05 



0.05 



—0.45 



120 



0.5 



59.5 



0.00 





0.66 



—0.66 



150 



1.7 



59.8 



0.75 



3 



1.27 



—0.52 



180 



1.1 



2.6 



1.85 



=0".61 



1.88 



—0.03 



