M. Bessel's Correction of the Thermometer. 307 



in many native oxides of iron. — A Report was received from 

 the Commission relative to Gas Illumination and Gasometers. 

 Its consideration was postponed to 



Feb. 9, when it took place accordingly. 



Feb. 16. — Arnaud Reynaud announced the discovery of a 

 method of protecting the Magnetic Needle from the influence 

 of iron. — M. Tilerier requested that a Report might be made 

 on his mode of making elliptic parabolic Mirrors. — M. Da- 

 moiseau presented a Memoir on the Perturbations of the Mo- 

 tion of the Comet of 1819 in the two periods which preceded 

 its perihelion passage in 1825. — M. Arago deposited in the 

 Archives the astronomical observations made at Paramatta in 

 June 1823, received from Sir Thomas Brisbane. — M. GeofFroy 

 presented a Table of corresponding Nomenclature of the sec- 

 tions of the Skull of various vertebrated animals. — The Commis- 

 sion on Gas Illumination presented some new propositions re- 

 lative to Gasometers placed at a distance from the Gas-works. 



LII. Intelligence and Miscellaneous Articles. 



bessel's method of ascertaining the corrections of 

 the readings of thermometers. 

 A CORRECT thermometer being a necessary instrument 

 ^*- in every observatory, we consider it useful to draw the 

 attention of the astronomers of this country to the imperfec- 

 tion of many of them, arising from the defective form of the 

 tubes. With this view, we lay before our readers the detail- 

 ed account of M. Bessel's method of correcting his thermo- 

 meter, contained in the seventh section of his " Astronomical 

 Observations," a book of which there are very few copies in 

 this country, and written in a language not generally under- 

 stood by our scientific men. 



" There are, perhaps, no thermometer tubes, the insides of 

 which are perfectly cylindrical ; all those, at least, which were 

 given me as such, proved on examination to be sensibly defec- 

 tive. Either a scale of unequal divisions, or a table of correc- 

 tions, is therefore necessary to derive accurate results from ther- 

 mometers constructed of such tubes. The latter deserves the 

 preference, as it is always easier to ascertain the errors of in- 

 struments than to make them perfectly correct. Let the 

 correction for any point x of the scale = $ x;^ this quantity 

 must be so determined, that for every column of mercury con- 

 tained between the points x and x of the scale, the quantity 

 (^4-<p x') — (x + f x) shall be a constant quantity, to whatever 

 part of the tube it may be moved. This being obtained, we 

 nave the following proposition : 



Qq2 5+<J>5 



