170 Introduction to the Seventh Sect ion 



constantly increasing under the patronage of the most liberal 

 Governments of Europe, and particularly of our own. 



We have been informed that Capt. Sabine has been lately 

 determining the length of the pendulum at various places on 

 the earth's surface ; which, from their accuracy and import- 

 ance, cannot fail to be highly interesting in the inquiry to 

 which we have now been directing our attention. 



Edinburgh, July 18, 1823. 



XXVIII. Introduction to the Seventh Section of Bessel's 

 Astronomical Observations. 



[Continued from p. 109.] 

 4. Flexure of the Telescope. 

 r | 1 HE effect of this flexure on the readings of the circle has 

 -*■ been determined by two methods : the first is founded on 

 a comparison of the zenith distances of a JJrsceMinoris obtained 

 by reversing the instrument with those observed by reflexion 

 from the surface of water; the second seems to be still more 

 sure and advantageous, being deduced from a comparison of 

 the distances of northern and southern stars measured both 

 by direct vision and by reflexion, and therefore independent of 

 the reversion of the instrument. Both methods are indepen- 

 dent of the divisions of the instrument ; the one position of the 

 instrument being in the same relation to the. zenith as the 

 other to the nadir; so that the direct observations in the one 

 position, and those obtained by reflexion in the other, depend 

 on divisions of the circle which are 180° distant from one an- 

 other ; and consequently are in both cases the same, the ver- 

 niers being likewise diametrically opposite one another. In 

 order, therefore, to eliminate the possible errors of division, 

 it will be proper to compare the direct observations in one 

 position of the instrument with those obtained by reflexion in 

 the other. 



These observations have some difficulty, because the slightest 

 draught of air will disturb the level ofthe water, and many 

 are therefore lost ; they rarely succeed when the wind is from 

 the side of the star that is to be observed ; and consequently 

 the distances of northern and southern stars are rarely ob- 

 tained by reflexion from water on the same day. This dif- 

 ficulty may be avoided by measuring the distances of the re- 

 flected images of stars from a fixed point in the heavens, and 

 by deducing the distances of the stars from the sums of such 

 observed distances. 



The fixed point which I have used, not only in these but 



likewise 



