on the presentation of a Medal. 



133 



surpass. But this was rather because, in the best observations from 

 Bradley's time forward, 1" has been a tolerated error ; a quantity for 

 which observation and mechanism, joined to atmospheric fluctuations 

 and uncertainties of reduction, could not be held rigidly accountable 

 even in mean results ; than from any reason in the nature of the case, 

 or any distinct perception of its reality. If parallax were to be de- 

 tected at all by observations of the absolute places of the stars, it 

 could only emerge as a " residual phenomenon," after clearing away all 

 the effects of the uranographical corrections as well as of refraction, 

 when it would remain mixed up with whatever uncertainties might 

 remain as to the coefficients of the former, with the casual irregula- 

 rities of the latter, and with all the forms of instrumental and obser- 

 vational error. Now these have hitherto proved sufficiently even in the 

 observation of zenith stars, quite to overlay and conceal that minute 

 quantity of which astronomers were in search. 



It is not my intention, Gentlemen, to enter minutely into the his- 

 tory of the attempts of various astronomers on this problem, whether 

 by the discussion of observations of one star, or by the combination 

 of those of pairs of stars opposite in right ascension ; nor with the 

 occasional gleams of apparent success which, however, have always 

 proved illusory, which have attended these attempts. For such a 

 history, and, indeed, for a complete and admirably drawn-up mono- 

 graph of the whole subject, I must refer to a paper lately read to 

 this Society by Mr. Main, and which is now in process of publication 

 in the forthcoming volume of our Memoirs. In whatever reference I 

 may have to make to the history of the subject, I must take this 

 opportunity to acknowledge my obligations to the author of this 

 paper, as well as for his exceedingly luminous exposition of the 

 results of those more successful attempts on the problem by Hender- 

 son, Struve, and Bessel, which I shall now proceed more especially 

 to consider. 



It would be wrong, however, not to notice that the first indication 

 of some degree of impression beginning to be made on the problem 

 seems to be found in Struve's discussion of the differences of right 

 ascension of circompolar stars in 1819, 20, and 2L The only posi- 

 tive result, indeed, of these observations is, that in the case of twen- 

 ty-seven stars examined, none has a parallax amounting to half a 



