340 AZIMUTH A L CHANGES OF HILLS, 



The results employed by me in my present investigation, 

 were obtained from the published Greenwich Observations for 

 the years 1858, 1859, 1860, and 1861. But, before I proceed, 

 I will briefly draw your attention to the mode of obtaining those 

 results. 



In a public observatory, the principle astronomical work 

 consists in observations of stars, as they are in the act of passing 

 across the meridian of the observatory. For this purpose an 

 instrument, called a Transit Circle, is employed, the application 

 of which, is sufficiently illustrated by the common theodolite, 

 bearing in mind that the axis is supported by solid stone piers, 

 sunk several feet into the ground, and that the telescope itself, is 

 not supposed to have any motion, except in a vertical plane 

 coinciding with the meridian of the place. Now if the instru- 

 ment were put in complete adjustment, and could be so retained, 

 the axis would be perfectly horizontal, and the line of collimation 

 of the telescope would always move in the meridian of the place ; 

 but as from mechanical defects and other causes, it is practically 

 impossible to ensure these conditions for any length of time, it 

 is usual, after the instrument has once fairly settled down, to 

 bring the errors of level and azimuth, as they are termed, 

 within moderate limits, and to determine, by very delicate ob- 

 servations made every evening, if possible, their actual amounts 

 at the time, and to apply the results thus obtained as corrections 

 to the ordinary star observations ; at the present time it is the 

 error of azimuth, with which we are principally concerned. 



If there were any fixed star situated exactly in the pro- 

 longation of the earth's axis, we should have nothing to do but 

 to turn the telescope upon that star ; and if, after perfect adjust- 

 ment in other respects, this star appeared upon a vertical wire 

 in the focus of the object glass, we should know that there was 

 no error of azimuth. And even if there is not this coincidence, 

 the error of azimuth, which ought always to continue small in a 

 good instrument, can be readily determined by means of a 

 micrometer. 



It is true that in neither Hemisphere is there a star exactly 

 coinciding with the corresponding Pole; but in the Northern 

 Hemisphere there are two stars " Polaris, of the second Magni- 



