of Vessel's Astronomical Observations. 343 



the use of my refractions, if the observations were made and 

 read off with absolute correctness. I find it as follows : 

 Z. D. Z. D. 



45° 0"-27 81° l"-00 



60 -34 82 1-11 



65 -37 83 1 -25 



70 -46 84 1 -43 



75 -66 85 1 -71 



80 -92 86 2 -40 



The errors originating in the dispersion of light, indistinct- 

 ness of the stars, &c, are comprehended in these quantities ; 

 in the vicinity of the horizon these disturbing causes act so' 

 strongly that one might be inclined to ascribe to them only 

 errors like those here given : I mention this as an additional 

 proof of the necessity of employing the factor A ; were it neg- 

 lected, and yet the tables used in such different temperature 

 as the winters and summers present at Konigsburg, one would 

 find much larger probable errors. 



8. Latitude of the Observatory. 



The latitude cleared from flexure, but still uncorrected for 

 errors of division and refraction, is by art. 4, =54° 42' 50' '785 ; 

 the errors of division in the eastern and western position of 

 the circle have been above determined to be respectively 

 = +0"-177and + 0"-325, from which follows the correction 

 — 0'*074: the new determination of the refraction produces 

 another correction of — 0"-188; so that the latitude re- 

 sulting from all observations with the meridian circle is = 

 54° 42' 5()"-52. 



Cary's circle formerly gave by zenith distances of a. Ursa; 

 Minoris ( = 36°) measured on both sides of the meridian 

 54° 42' 50"-276-0"'452A8, or putting the correction of the 

 declination taken from the tables = — 0"'339, 54° 42' 50"'43, 

 from which is to be deducted for the new correction of re- 

 fraction 0''*19. Between these two determinations there is 

 therefore a difference of 0"-28, which is indeed a little beyond 

 the limits of the probable error, but affords no sufficient rea- 

 son for supposing that any constant error has been neglected. 

 Which of the two determinations deserves the preference, I 

 do not dare to decide, as I consider it to be exceedingly diffi- 

 cult to arrive at certainty with regard to such small quantities. 



9. Solstices of 1820 and 1821. 



The solstices having frequently given rise to doubts respect- 

 ing the accuracy of the declinations, I think that I may insert 

 licit: what my observations and the methods of reducing them 

 Ihic explained have broughl out. 



