respecting the new Tables of Refraction. 263 



out any adventitious advantages ? I have therefore computed, 

 by the table in the Nautical Almanack, the same refractions 

 which are given in my paper, p. 490, as calculated both by 

 the French table and by my own ; and, reckoning the excess 

 of the calculated above the observed quantities for the error, 

 the results of all the three tables, when they are arranged in 

 the order of the stars, will be as below : 



N.A. F.T. N.T. 



These calculations are easily verified. The same observa- 

 tions are discussed in a former Number (No. 29) of the Journal 

 of Science. The point aimed is to make the tables approach 

 as near as possible to the observed quantities. This is effected 

 partly by shifting the mean temperature from 50°, which is 

 the standard degree assumed in the construction of the table, 

 to 46i°; and partly by altering the allowance made for 

 one degree of the thermometer in the corrections for heat. 

 Thus, excepting what regards the barometer, the table is en- 

 tirely altered as far as the corrections are concerned. It is 

 therefore highly ridiculous to say that the results in the Jour- 

 nal are calculated by the table in the Nautical Almanack. In 

 the case of Mr. Groombridge's observations, when the com- 

 putations are fairly made, the errors of the table in N. A. are 

 much greater than those of either of the other two tables. 

 And, I apprehend, it will be found, on examination, that the 

 table in N. A. is less accurate generally at low altitudes than 

 the French table. 



Besides the observations just mentioned, the results of nine 

 others are given in the Journal of Science ; but of these I 



shall 



