The Rev. Dr, Robinson on the Constant of Refraction. 189 



either by direct experiment, as was done by Arago and Blot,* or by astronomical 

 observations. Notwithstanding the well known accuracy of these distinguished 

 philosophers, it seems desirable that their conclusions should be verified by the 

 more refined means of examination, which Arago himself has since indicated. 

 At present, the result appears in excess, giving for fi at 50° and 29'.60 the value 

 57". 82. That which is most generally received is De Lambre's, employed in 

 the French tables, as well as in those of Brinkley and Ivory. It is at the same 

 temperature and pressure 57". 72, and was deduced from observations made with 

 the repeating circles of Le Noir, so that it would not have much weight now 

 were it not for the confirmation which it seemed to derive from the comparison 

 of simultaneous observations by Brinkley and Brisbane, at Dublin and Paramatta. 

 The sum of the Dublin north polar, and Paramatta south polar distances gives 

 very nearly 180 degrees, and the resulting value of /x is 57-77 ; but it must be 

 remarked, that the temperature used in computation is that by the internal 

 thermometer, which, however necessary at Dublin, may not be so at the other 

 observatory. It is also important to notice, that the Dublin barometer is by no 

 means perfect. I have been enabled to determine its error by comparison with 

 that of the Magnetic Observatory of Trinity College, (by Newman, and differing 

 from mine and the standard of the Royal Society merely in having the cistern of 

 glass.) Observations made during thirteen successive days at 22 " give 



A. T. 



41.60 

 37.70 



The difference of height of these stations is, according to Captain Larcom, 258.8 

 feet, and I compute that the actual pressure at the upper station was 29.702 ; so 

 that the reading there requires the correction -|- 0.077. Subsequently this has 

 been confirmed by the kindness of Dr. Coulter, who compared two portable 

 barometers, by Cary, with that of the magnetic observatory, very carefully. They 

 were then carried out to the astronomical observatory, compared there, and on 

 their return compared again with the magnetic. From the result of the two 

 sets I deduce the corrections -\- 0.0770, and + 0.0800, the mean -|- 0.0785 

 I consider preferable to the other, and this would reduce the constant 57.72 to 



* Memoires des Scavans Etrangers, T. vii. 



