of BessePs Astiyntmical Observations. 



433 



The probable error of a levelling of the axis is, therefore, 

 0'112 line =0"*243, and consequently that of the above mean 

 = 0*0201 line ; from which it is clear, that there cannot be any 

 doubt of the above-mentioned irregularity. The small devia- 

 tion probably does not proceed from an imperfection in the 

 cylindrical figure of the pivots, but from their axes not being 



Vol. 63. No. 314. June 1824. 3 I exactly 



