8G On the Royal Observatory at Palermo. 



From 1755 to 1798 = — 0",256 

 1798 to 1806 = + ,188 

 1806 to 1812 = - 0,767 

 1812 to 1824 = + ,800 

 Take another instance in tlie case of Y^Eridani: the proper 

 motion in dechnation appears to be as follows : viz. 

 From 1755 to 1798 = + 0",135 

 1798 to 1806 = - ,125 

 1806 to 1810 = + 1 ,500 

 1810 to 1812 = + 1 ,250 

 1812 to 1824. = + ,050 

 1824 to 1825 = + 1 ,000 

 Several other similar instances might be adduced of an ap- 

 parently variable proper motion : we shall select however only 

 one more, which seems to be the most remarkable. It is the 

 case of 67 Urs(V Majoris: the proper motion of which appears 



This subject is certainly a very curious and interesting one, 

 and is deserving of more minute inquiry and investigation : 

 but we very much doubt whether the observations of Bradley 

 or Piazzi (setting aside those of Mayer, La Caille, &c.) were 

 made with sufficient accuracy to serve as the foundation for 

 such a minute investigation. These remarks are not made 

 with a view to diminish the fame of those excellent observers, 

 whose characters stand too high to be thus affected ; but when 

 we see that the best practical astronomers of the present day, 

 with the most perfect instruments that ever were made (far 

 superior to any that could have been possessed by their pre- 

 decessors), differ from each other, not only one or two, but 

 sometimes as much as four and five seconds, in the position of 

 a star, both in right ascension and declination, we may fairly 

 hesitate in assenting to the conclusions which are intended to 

 be inferred from these comparisons. There are moreover too 

 many discordant elements in the reduction of observations, too 

 many circumstances to be attended to in the steps of that pro- 

 cess, to induce us to assent instantcr to the results which are 

 thus laid before us. The detail of the method pursued in such 

 cases cannot be too explicit, where the differences are so mi- 

 nute, and the results so important. It is not enough to com- 

 pare one reduced catalogue with another reduced catalogue, 



unless 



