5b 



Astronomical Science, §c. 



[JULT 



On examining the column " Mean" for the right ascension, it will 

 probably be remarked that the ten thousandth part of a second of time, 

 or even the thousandth part, is a quantity over which our observations 

 can possibly exercise no control. In anticipation of such a remark,. 

 I have computed, that the probable error of each of the above values 

 =,0006 and the most probable largest amount of error =,0023*. If we 

 now re-examine the column " Mean" for right ascension, recollecting 

 that any or each result may be erroneous to the amount, 0023, we find 

 that something remains to be accounted for. On examining the column 

 " mean" for declination, a general determination towards the minus 

 sign shews that some error exists in the assumed latitude, either of 

 Palermo or Madras, or in both, to the amount of 1,46*. Here, too, it 

 might be remarked, that the observations cannot possibly take cogniz- 

 ance of the ten thousandth part of a second of space, — on computation, 

 I find the most probable largest amount of error =,"0102. If we now 

 reduce the column " mean" by the correction ,"0417 on account of the 

 error of latitude, we obtain the column " corrected mean" Examining 

 this column, recollecting that errors to the amount ,"0102. may exist in 

 any of the results, we again find, as in the case of the right ascension 

 column, that something remains to be accounted for. 



It would not be wise here, to endeavour to reduce the numbers in 

 either column to regularity (by attributing to them errors within the 

 above found limits), and to proceed upon such regularity to establish 

 the law of change, and then to shew that such law can be explained by 

 such and such a cause, &c. No — the plan is, to continue the observa- 

 tions to a larger number of stars, and allow the observations them- 

 selves to point out the law, in a manner which will not admit of 

 doubt. 



The above results are derived from the observation of 3,003 stars, 

 during the years 1834 and 1835, in addition to which, I shall shortly 

 (in a year and a half from the present time) be prepared to encounter 

 the enquiry, with a further catalogue of five thousand stars, when the 

 cause of this now seeming anomaly will, I have no doubt, be satisfac- 

 torily explained. 



