514 Proceedings of the Royal Irish Academy. 



observing (that is, using the old or the new method), and that the 

 appearing differences were a good deal larger than the probable errors 

 of a single observation. Before deducing final results with respect to 

 the positions of northern stars, it seemed therefore necessary directly 

 to find the above-mentioned differences between the different princi- 

 pal observatories ; or better, to determine the absolute personal errors 

 by one of the artificial instruments. For the first reason, M. Wagner 

 went in the summer of 1868 to Greenwich, where he, on seven nights, 

 compared himself with the transit observers there ; for the second 

 purpose, a time-collimator was placed in one of the meridian-mark 

 houses in Pulkowa, so that the transits of the artificial star could be 

 observed by lenses of proper focal length, in the large transit instru- 

 ment. 



Polar stars are in Grreenwich still observed with eye and ear 

 alone. As the observations are taken by several observers, M. 

 Wagner could only obtain mean results. While his equation with 

 the Greenwich observers for equatorial stars was very small,'^' he 

 found for stars near the pole : 



s\ p. D. 



"W. - Greenwicli. 



Number 

 of Stars. 



5° 12' 



- 0^-02 



3 



4 22 



-0-51 



3 



3 24 



-0-02 



7 



2 45 



-0-50 



6 



1 5 



-0-34 



1 



1 



- 1-14 



5 



The negative sign and the increase towards the pole seems certain. 

 This result became more interesting, when it appeared that some of 

 the observers differed more than others. By adopting the expression 



sin 2°4 5' 

 ""^ siniVPj 



the following differences were found : 



W.-Dunkin=- 0=-56 



W. -Ellis =-0-17 



sin 2°45' 

 siniVi^ 



sin 2°45' 



siniVPi)' 

 the probable errors of the coefficients being respectively = + 0'"15, 



* Grreenwich Observations, 186S,Intro(i. (Dunldn- "Wagner) = - O'OSbychrono- 

 graphic observations. 



