386 



Motion of the Solar System in Space. 



On examining the above table, we notice that the sign minus inva- 

 riably prevails, which leads us at once to enquire if any correction 

 for a wrong assumption of I, or I', the latitudes of Madras or Palermo, 

 are necessary; and how far the use of an incorrect table of refraction 

 may affect the question ; added to which we have already found that 

 a correction of + 0,04'' is necessary to the general precession in 

 longitude which has been employed ; rendering necessary, to each 

 group in the above table, the correction ,015 cos a, or the correction 



-J 015 cos a : now the value of -j depending upon t, the number 



of years since 1800, will necessarily be larger for the catalogue of 

 vol. II. (for 1832), than for those of vols. in. or iv. 5 whereas the re- 

 maining part of the correction varying with «, the right ascension of 

 the group, will be constant for each catalogue at the same hour, but 

 ■will vary to the same extent -f as — ; hence, the mean of the 24 

 results for each catalogue, exhibits (independent of any proper motion 

 which may exist) the values of j-thus, on taking the means, we get 



corrected mean proper motion in declination = — , 0544 -}- 



whence S — 3.''6i. 



Considering the general effect of the worst tables of refraction yet 

 published, upon a group of stars, such as is met with in each hour of 

 A. R., the error in the Palermo results or d r, cannot err beyond half a 

 second of space ; whereas in the Madras results, in which the stars are 

 evenly distributed at either side of the zenith, the uncertainty does 

 not amount to one-tenth of a second. "With regard to the latitude of 

 Palermo, it may be wrong, for we have no evidence to shew its ac- 

 curacy or inaccuracy ; whereas the latitude of Madras from recent 

 observations should be diminished 1"; we have altogether as follows, 

 S = 3/61=-. dr—V ,0 +0,5+, 0 .-. d is between 4,"1 and 5;'i 



Now an error of above 4 - in the latitude of Palermo, were it only 

 out of respect to its late distinguished Astronomer, must not be ad- 

 mitted upon slight grounds ; we will therefore subtract the mean of 

 each catalogue from each of its constituents' values, and then combine 

 the results according to their respective weights ; thus, putting * for 

 the correction which may now remain due to render them just. 



altogether = d. I + d. r + d. r + d. r ' 



-f , 015 cos a, which put = 



