Variations of the Proper Motions ofProcyon and Sirius. 257 



As the right ascensions of Sirius, which are given for the be- 

 ginning of each year in the TabultB Regiomontance, are obtained by 

 comparison of the right ascension in 1755 with that in 1825, their 

 agreement with the observations of the latter years was complete ; 

 but, as early as the year 1835, fifty observations showed, when com- 

 pared with the three fundamental stars following Sirius, viz. /3 and 

 a Orionis and a Canis Minoris, that 0*'188 must be added to the 

 Tabulae Regiomontana, to make the agreement again perfect. This 

 disagreement has Ijeen since still increasing. In the year 1843 I 

 found it, from fifty observations by Dr. Busch with the old instru- 

 ment, = + 0**318, and, from forty, which I made myself with the 

 new meridian circle of Repsold, I found it = -}- 0*'324. 



A second suspicion of the variability of the proper motions of the 

 stars was awakened in me in the year 1840, by the declination of 

 Procyon, since a new determination of all the elements of reduction 

 of its declination, and of the declinations of the other fundamental 

 stars, gave the observed declination of Procyon more northerly than 

 that of the Tabulce RegiomontarKS (the tabular result being obtained 

 by comparison of the observations of 1755 with those of 1820), by 

 1"'64*. lliis difference has also increased, since, by observations 

 made with the instrument of Repsold, I find it for 1844 = + 3"* 18. 



What I have brought forward concerning Sirius and Procyon de- 

 pends on determinations, whose certainty I esteem as great as can 

 be attained by the present apparatus for observing. At the same 

 time it does not cease to be necessary to subject the important re- 

 sult here given to the strictest scrutiny, by means of all the existing 

 determinations of other observatories, before it can be received as 

 the indisputable result of observation. I would communicate the 

 result which this investigation has produced ; but I should go be- 

 yond the limits of a letter, should I here give place to the criticism 

 to which some of the numbers must be subjected, before they can be 

 received as valid. Since this is, nevertheless, not the less neces- 

 sary, I must refer to a paper which will very soon appear on this 

 subject in the Astronomische Nachrichten. It is plain that we can 

 obtain for the declination of Procyon comparative results from the 

 different catalogues, only by eliminating the constant errors, which, 

 without doubt, affect all the observations up to the present time, 

 and frequently to the amount of several seconds. This I found to be 

 the case by subtracting from the difference between every deter- 

 mination of Procyon and the Tabulce Regiomontance, the mean of 

 the differences of the results for eight stars, a, Ceti, a Orionis, 

 /3 Virginis, a Serpentis, y, a, /3 Aquilse, and a Aquarii, the mean of 

 whose declinations is the same as that of Procyon, within a very 

 few minutes. By this means it will be gathered that the following 

 collection of results does not depend upon the absolute declination 

 of this star and the absolute right ascension of Sirius, but upon the 

 relative declination and right ascension of each respectively, as 

 founded on the comparison with the above-mentioned eight and 

 three stars. 



* Ast. Nachr. No. 422. 



