Declinometer. xiii 



meter, is divided to 5 minutes, and is read to 5 seconds with the three verniers. 

 By some accident, the circle has probably been flattened on one side, as there the 

 verniers enter rather too much on the graduations ; the error due to this is, how- 

 ever, very small. The lines of coUimation of the theodolite and reading teles- 

 copes coincide when the middle wire of the former is made to coincide with the ver- 

 tical wire of the latter. The circle is retained in the same position on its pillar, 

 but the transit telescope is removed, excepting when required for observations of 

 absolute declination. The reading telescope is fixed to its pillar. Following are 

 the determinations of the data requisite in reducing the observations of the decli- 

 nometer. 



8. Value of the declinometer scale divisions in angular measure. 



The scale generally used consists of 500 divisions ; but, during observations 

 of absolute horizontal intensity, the magnet with this scale is removed to the uni- 

 filar magnetometer, and a magnet with a scale of 300 divisions is substituted in 

 the declinometer. The observations. Table 2, Introduction, for 1841-2, gave one 

 division of the long scale = 0"'6710. Observations were also made on December 2, 

 4, and 5, 1843, and on October 22, 1844. The results were as follow : — 



Dec. 2. 1843. Two series of observations, only one vernier of the circle read. 



1st series ; measures of 200 divisions ; one division of the long scale = 0'-6725 

 2d, =0'-6725 



These results were obtained after applying a correction on account of the non-coin- 

 cidence of the graduations of the vernier with those of the circle. 



Dec. 4. 1843. One series ; measures of 100 divisions ; one vernier read ; one 

 division of the long scale =0*'6728. 



Dec. 5. 1843. Before the previous observation, it was found that the lens of 

 the collimator was somewhat loose j it was then, it is believed, made more so. This 

 was owing to the frame not being screwed hard up into the collar, which was now 

 done. 



One series ; one vernier read ; measures of 100 divisions ; one division of the long scale 



rC 6715 1 , fB. 



= l0'-672l] ' °^^^^^^^ [w. 



Oct. 22. 1844. Owing to the difference of the results on Dec. 5. 1843, from 

 those on the previous occasions, the following short series of observations were 

 made with much care. All the three verniers were read, the readings being made 

 first with the right and then with the left eye ; the mean of the two being taken as 

 the true reading. 



MAG. AND MET. OBS. 1843. d 



