360 Ordnance Survey Astronomical Observations. 



it may do better ; though there is even then great doubt if 

 the level ever can be ascertained as exactly as it should be ; 

 and after all, eacli observation occupies so very great a length 

 of time, as to admit of but very few being taken during a 

 night ; and those may be so seriously interfered with by the 

 abundant clouds of this climate, that a practical man may 

 well prefer a graduated circle in the Meridian. 



Second, — Of the mode of observing, and reducing the ob- 

 servations. 



Each observation before us is made to give an independent 

 latitude, and this is therefore loaded, not only with the in- 

 strumental errors in measuring the zenith distances, but also 

 with the tabular errors of the star's declination, and what is 

 much more uncertain, its proper motion. The prejudicial 

 consequence of this appears in the columns on the right hand 

 pages giving the observed zenith distances and the computed 

 latitudes, in which latter the discordances are often double 

 the amount of the former. 



Uncertainties of refraction are certainly mixed up in the 

 latitude results ; but as the zenith distance is always small, 

 there can never be any notable effect from that cause ; yet 

 we should like, in any future work, to see a discussion of the 

 value of the constant of refraction ; for it would doubtless be 

 found to vary at the different altitudes above the sea, and in 

 the contrasting physical circumstances of the several stations. 



The greater part, then, of the increase of the discordances 

 amongst the latitudes over the zenith distances, must rest 

 with errors in the assumed values of the places and proper 

 motions of the stars of comparison ; and an easy way of 

 remedying this difficulty is to have two instruments : keeping 

 one always at a standard station, and having the same stars 

 observed at the same times, with the stationary and with the 

 travelling instrument : a method much practised by Mr Mac- 

 lear in his excellent repetition of the South African arc. 



Thirdly, we may observe, that while we should prefer, at 

 the beginning of the volume, to see entered the very identical 

 numbers read off from the instrument, without, as here, being 

 reduced from divisions of the micrometer to seconds of space ; 

 we should also like to see at the close, some investigation 



