OF THE COSMOS. PABALLAX OF STARS. 187 



pendent on the improvement of optical and measuring in- 

 struments, and on the possibility of determining very small 

 angles with certainty. So long as such certainty was only 

 equal to one minute, the non-detection of parallax only proved 

 that the fixed stars must be more distant than 3438 semi-dia- 

 meters of the Earth's orbit. ( 306 ) Certainty to a second in 

 the observations of the great Astronomer, James Bradley, 

 raised this lower limit to 206265 semi-diameters; and in the 

 brilliant epoch of the Fraunhofer instruments, it was raised, 

 by the direct measurement of nearly the tenth part of a 

 second of arc, to 2062648 semi-diameters of the Earth's 

 orbit. The efforts, and ingeniously devised Zenith apparatus, 

 of Newton's great cotemporary, Robert Hooke, in 1669, did 

 not conduct to the desired aim. Picard, Horrebow who 

 worked out Homer's rescued observations, and Elamsteed 

 thought they had found parallaxes of several seconds, be- 

 cause they confounded the proper motions of the Stars with 

 the effects of parallax. On the other hand, the sagacious 

 John Michell (Phil. Trans. 1767, Vol. Ivii. pp. 234-264), 

 was of opinion that the parallaxes of the nearest fixed stars 

 must be less than /x '02, and therefore " could only be recog- 

 nised by a magnifying power of 12000." From the very pre- 

 valent idea that the superior brightness of a star must always 

 indicate its greater proximity, stars of the first magnitude, 

 Yega, Aldebaran, Sirius, and Procyon, were first studied, 

 and were the subjects of the unsuccessful observations of 

 Calandrelli, and of the meritorious Piazzi. (1805) To these 

 observations, we must add those which were published (1815) 

 in Dublin by Brinkley, and which, ten years later, were re- 

 futed by Pond, and especially by Airy. A sure and satis- 

 factory knowledge of parallaxes, founded on micrometric 



