368 



Mr. S. Lee on some Inconsistencies 



consistencies, will evidently appear, from the following short 

 abstract of a comparison of the observations as recorded in 

 different parts of the book with each other ; and which, for 

 the sake of perspicuity, I have arranged in separate classes. 



The first class are those which were pointed out to me by 

 my friend, and are to be found in the North Polar Distances, 

 deduced from the observations with the mural circle, from the 

 readings with two — and with six microscopes. Those discover- 

 ed in the observations with two microscopes alone are at least 

 sixty in number, and frequently amount to more than 1". 



The second class was discovered by comparing the read- 

 ings with the microscopes A and B, as given under the head 

 of observations with two microscopes, with the same given un- 

 der the head of observations with six microscopes : of these 

 errors sixty-four have been discovered. 



The third class are to be met with in the extraordinary dis- 

 agreement between the heights of the barometer and thermo- 

 meters, as recorded under the separate heads of observations 

 with two — and with six microscopes ; and amount to two hun- 

 dred and eighty-six in number. 



I shall content myself at present with a bare statement of 

 these three classes of errors only, reserving any observations 

 on them, and some others, of perhaps still greater consequence 

 to the reputation of the observatory, to a future communication. 

 In the mean time I think it is due to the character of Mr. Bens- 

 ley, to state, that they are not errors of the press, a fact which 

 I have ascertained by a comparison of the printed copy with 

 the original manuscript deposited in the archives of the Royal 

 Society. I am Gentlemen, 



Your obedient humble servant, 

 London, Nov. 24, 1824. Stephen Lee. 



Class I. 



Page. 



177 9 Feb. 

 185 <$ Apr. 



9 20 



S June 26 

 h Aug. 11 

 d Oct. 29 



186 

 194 

 200 

 209 

 210 ... . 



O Nov. 4 

 219 % Dec. 13 



And other similar 



True 

 Mean. 



13-8 

 10-6 

 >r 40-6 

 6-5 

 16-1 

 52-8 

 21-5 

 ll-l 

 505 

 44-0 



errors, amounting in all to 60 in number. 



