The Rev. Dr. Robinson on the Constant of Refraction. 211 



The value of ft used in computing the refractions is, 



/x = 57.7682 ; 

 <//! = — 0.2255 ; 



sum = 57.5427. 



This may perhaps require a correction for the run of the microscopes, which 



though very small is sensible. From the erection of the circle to July 8, 1837» 



0" 18 ">( a' 

 its effect on the mean of four microscopes was = — — '~r77r — ^* *^^^ *^™^ 



it was changed by the rough operations necessary in attaching another pair of 



a' 

 microscopes, and has been since considered permanent at -f- 0". 41 X —;• This 



is, however, a mean value, being deduced from readings of the four, in 30 

 equidistant positions of the circle. Hence 1 found as above 



_38^2909^ 

 ^ ^10225.547 ^ 

 and 



M = 57".5464 



a value whose near approximation to Bessel's 57".524, will prove very remark- 

 able, if when I have means of determining the length of the seconds' pendulum 

 here, it should be found little different from that of Konigsberg. That obser- 

 vatory is a little north of me, but it is only 90 feet above the Baltic ; while this 

 is 211 feet above the sea, and the substratum, dense limestone, so that the local 

 gravity must be nearly alike in both cases. 



As to the southern stars, I have used the declinations of the St. Helena 

 catalogue, reduced to Bessel's refractions, by the table given page 22, and 

 those of Professor Henderson. (Mem. R. Ast. Soc. X. 80.) The two are not 

 strictly comparable in respect of refraction, for the St. Helena Observatory, 

 being 700 feet above the sea, and resting on dense volcanic rocks, may be 

 expected to have an excess of gravity above the Cape, and therefore larger 

 refraction. At the latter place I find, by comparing the length of the pen- 

 dulum with that of Greenwich, that Bessel's refractions should be multiplied by 

 0.9984 ; and, in fact, Henderson's observations on refraction shew, that even a 

 greater diminution is required. I have not, however, changed them further than 



2 E 2 



