VOL. LXXVII.j PHILOSOPHICAL TRANSACTIONS. 223 



use of the true parallax 8'''8 or 8^''', as found by the two late transits of Venus 

 over the sun, he would have made the refraction at the altitude of 4S° to be 56^'^ 

 instead of 57^ and the latitude of the observatory exactly 51° 28' 40'''' instead of 

 51° 28' Sg^". But his rule for refractions cannot be corrected for all altitudes, 

 without examining his observations of refractions made at various times." 



On comparing this extract with M. Cassini's memoir, I cannot but express my 

 surprise, that he sh9uld not have adverted to a passage containing so direct an 

 application to the grounds of his memorial, in a publication of such notoriety, 

 and of so old a date as 177^; had he done so, I cannot but think he would 

 never have hazarded such an opinion as that advanced by him in his memoir, of 

 an uncertainty of 15''' in the latitude of Greenwich; but he might have been 

 induced to believe, that the latitude of this place had been well determined. 



For further confirmation of the certainty of the astronomical refractions, and 

 latitude of the observatory, as settled by Dr. Bradley, it may be proper to add, 

 that the Greenwich brass mural quadrant underwent a trial, which all astrono- 

 mical instruments ought to be submitted to, but which very few ever have been, 

 on account of the difficulty and nicety of the operation, namely, an examination 

 of the total arc; when it was found by Dr. Bradley to be an accurate quadrant, 

 the arc appearing at one trial to differ only a fraction of a second from 90°, and 

 another time, after an interval of above 6 years, to be a perfect quadrant. See 

 p. 24 of Bird's Method of constructing Mural Quadrants, published by the 

 Board of Longitude in 17 68. In like manner he had before examined the total 

 arc of the iron quadrant, first put up by Mr. Graham, for the uSe of Dr. Halley, 

 in the year 1725, by means of a level, and found it to be l6'^ less than a qua- 

 drant. See Bird's Method of constructing Mural Quadrants, p. 7, and Me - 

 moires of the Royal Academy of Sciences at Paris, for 1752, p. 424. But this 

 quadrant was, in the year 1753, re-divided by Mr. Bird, and in this respect pro- 

 bably rendered as accurate as the other. See Bird's Method of constructing 

 Mural Quadrants, p. 24. 



Dr. Bradley made a curious use of the new set of divisions, soon after they 

 were laid on the quadrant, to re-examine the error of the total arc laid down 

 originally by Mr. Graham (which by the plumb-line and level he had found to be 

 16" less than a quadrant in 1745) according to the following passage contained 

 in the manuscript before cited. *' August 12, 175-3, I measured with the screw 

 of my micrometer the difference of the arcs (of -§4) as set off by Mr. Graham 

 originally, and by Mr. Bird when he put on a new set of divisions on the old 

 quadrant, and I found that Mr. Graham's arc was less than Mr. Bird's by ^\ 

 divisions of my micrometer, which to a radius of 96 inches answers to 10'\6 ; so 

 that the whole arc of 96 differs from a true quadrant 15''.9, which is the same 

 difference that I formerly found by mean§ of the level, &c." 



