223 



On the other hand, when five double observations, in January and 

 February 1814, were added to these 61, they reduced the result for 

 the coefficient to 0"'18 ; so that the discordancies seem to be too great 

 to enable us to place any reliance on the conclusions respecting the 

 actual magnitude of the annual parallax. 



A similar fluctuation is observable in the results obtained for the 

 following years ; and though it might, on the whole, be inferred that 

 the parallax is about three fourths as great as that which the author 

 has assigned from his own observations, yet he contents himself with 

 poncluding that the mural circle of Greenwich has not sufficiently 

 proved the identity of the distance of the two stars in summer and 

 winter, within one tenth of a second ; but, on the contrary, that it 

 shows the parallax of a Lyra to be half a second greater than that 

 of y Draconis. 



In 1815 the first 15 summer observations, compared with the first 

 13 in November, give a parallax of + 0"'72 ; the next 16 in summer, 

 compared with the next 16 in winter, give a negative parallax of 

 0"'58 ; a comparison which sufficiently proves the imperfection of 

 the observations, depending probably on an unsteadiness in the in- 

 strument. 



In the whole five years, the mean of all the observations in August 

 exceeds the mean of July by 0"'51, a discordance which parallax 

 would diminish but in an inconsiderable degree. 



The author pursues a similar train of argument in the second part 

 of the inquiry, relating to the absolute parallax of a Lyrae. While 

 the circle at Dublin, he observes, made from a mean of several years 

 the double zenith distance of this star 3" greater in the beginning of 

 December than in the beginning of August, that of Greenwich shows 

 no difference whatever in the double altitude observed by reflection 

 in summer and winter. There are, however, differences of about 

 four seconds in the difference of altitude of a Lyrae and the pole star, 

 as determined in different years by the same instrument ; and Dr. 

 Brinkley observes, that an unsteadiness, amounting to 15" or 20", is 

 discoverable in the comparative results of the different microscopes ; 

 whence he infers that there must be an uncertainty, amounting to 

 many tenths of a second in the mean. 



The coefficients of aberration and of solar nutation, which come 

 out 20"'35 and 0"'51, are certainly true to one fourth or one tenth 

 of a second, as deduced from the observations of Dublin ; the author 

 thinks it fair, therefore, to infer that 1"'14, the coefficient for annual 

 parallax for a Lyrae, is correct nearly in the same proportion. Nor 

 are there any changes from season that could produce the appearance 

 of regular parallax of all the stars of which it has been inferred ; and 

 it is very improbable that any error of the instrument could have given 

 a parallax to a Lyrae, and left the pole star completely free from it. 



The last of the tables shows the consistency of the circle of Dublin 

 in the places of the stars, as determined by it after the interval of a 

 considerable number of years, without any such tendency to the south 

 as is supposed to have been observed at Greenwich. 



