14 



moveable vrire>. and the value of its scale was well determined both 

 by Mr. Beaumont and Lord \Vrotte5leY. The power usually em- 

 ployed was 450. 



In the progress of the observations it was found that they were 

 attended with considerable difficulties, chiefly arising from the im- 

 practicability of obtaining a sufficient number of observations at both 

 the proper periods of the year, and also from the circumstance that 

 many stars require to be observed at an inconvenient hour in the 

 early morning, when the observer, fatigued by night observing, is 

 unequal to the task. The fact of the difficulty of the observations is 

 evidenced by the paucity of the trustworthy results, after more than 

 six years' uninterrupted observing, viz. from February of 1843 to 

 October of 1S49. Of sixty-nine stars proposed for observation only 

 forty-eight have been observed, and of them nineteen only have been 

 observed at both periods of the year. The results of the obser- 

 vations of these nineteen stars are given in the paper. 



Four tables are given, the first of which contains the separate 

 results for each day's observations of every star, both for distance 

 and angle of position, with the probable error and weight of each, 

 computed according to the ordinary Ibrmula of the calculus of pro- 

 babilities, and also with the assigned arbitrary weight of each, esti- 

 mated according to the judgment of the observers. It contains 

 also estimations of the magnitudes and colours of the stars for each 

 night. 



The second table gives the similar results combined for each 

 period of observation, with tbe computedweights and probable errors. 



The third table gives the conibined results of all the observations 

 for the main epoch of observation, together with the approximate 

 R.A. and N'.P.D., and the whole number of observations. 



The fourth table gives the results for the separate epochs for those 

 stars only which afford reasonable hope of the detection of parallax, 

 four stars being omitted as evidently binary systems, and some others 

 whose components were equal in magnitude, and the observations of 

 which did not give any indication of parallax, being also omitted. 

 The differences of the angles of position, as indicting parallax, are 

 distinctly exhibited, first as resulting from such observations as were 

 made at consecutive and opposite seasons, and secondly as resulting 

 from the comparison of all the observations made at one period of 

 the different years with all made at the other period. 



In discussing the final results, the author remarks that only three 

 stars, viz. 118 Tauri, 100 Hercules, and Herschel 95, were observed 

 satisfactorily at the opposite and consecutive seasons, and these 

 exhibit such discordances in the partial differences, that it seems 

 necessary in all cases to depend only upon the average difference 

 of position obtained by comparing all the observations at one 

 period of the different years, with all at the other period. He finds 

 that there are then only five stars in the list, viz. 32 Eridani, 41 

 Aurigse, c Geminorum, an anonymous star in Cancer,and Herschel 95, 

 which show differences in the proper direction, and so much greater 

 than the probable errors as to deserve much attention, as exhibiting 

 parallaxes measurable by this method, and, of these, 3 Geminorum 



