THERMOMETER AND PYROMETER, 



55 



between the freezing and the boiling 

 point of water.* Petit and Dulong 

 have investigated the subject, by a com- 

 parison with the air thermometer, and 

 with the expansions of a pyrometer of 

 very infusible metals (platina and cop- 

 per) which appear, by the experiments 

 of Lavoisier and Laplace, to be very 

 equable in their expansions, below the 

 boiling point of water. They found 

 that the irregularity in mercury in- 

 creases with the temperature ; and 

 would even appear to be greater than 

 the rise of the mercurial thermometer 

 indicates, were not the increasing ex- 

 pansion of the mercury diminished by 

 the increasing ratio of the dilatation of 

 the glass itself, f Hence the source of 

 error in thermometers arising from the 

 expansion of the glass, is rather advan- 

 tageous than detrimental to their accu- 

 racy. 



The difference between the indica- 

 tions of the alcoholic and mercurial 

 thermometers, as lately ascertained by 

 Dr. De Wildt, do not materially differ 

 from the determinations of De Luc ; 

 and having been obtained for every 5 

 of Reaumur's scale, apparently with 

 much care, we give the result as a table 

 of correction for Rutherford's thermome- 

 ters ; for which purpose they were in- 

 tended by the author. 



Mercury. Spirit. Mercury. Spirit 



45 = _-28.50 -}-20...+16 .48 



40 ....25 .92 25 20 .97 



35 23 .19 30 25 .60 



30 20 .32 35 30 .38 



25 17 .30 40 35 .31 



20 14 .13 45 40 .38 



15 10 .82 50 45 .60 



10.... 7 .36 55 50 .97 



5 3 .75 60 56 .48 



.00 65 62 .14 



+5 + 3 .90 70 67 .95 



10 7 .95 75 73 .90 



15 12 .14 80 80J.OO 



The indications of air thermometers 

 were at one time supposed liable to un- 

 certainty, from the inequalities of their 

 expansion. Guyton and Prieur ima- 

 gined that they progressively expanded 

 in a greater ratio than the temperature, 

 but this has been proved to be errone- 

 ous ; and the mistake probably arose 



* Annales de Chimie, t. xliii. 



-j- Annales de Chimie et Physique, t. ii. p. 240. 



J Jameson's Edin. Phil. Journal, Oct, 1826, and 

 Kastner Archiv fur die Gesammte Natural, Decem 

 1825. 



$ Journal de 1'Ecole Polytechnique. 



from their neglecting the effect of hy- 

 grometric water in the gases. General 

 Roy* found that their expansion fol- 

 lowed a ratio, decreasing with the ele- 

 vation of temperature ; and the same 

 result was obtained by Mr. Dalton.-f- Dr. 

 Murray, $ Gay Lussac, and Petit and 

 Dulong ; || but there is reason to con- 

 clude that this apparently decreasing 

 ratio in the expansion of gases is owing 

 to the error caused by the unequable 

 expansions of the mercury in the ther- 

 mometer, and the dilatation of the bulb of 

 that instrument ; and philosophers now 

 agree to consider the expansions of 

 gases equable and equal, as before 

 stated; especially as the decreasing 

 ratio disappears, if we apply De Luc's 

 correction of the real mean between 

 32 and 212. 



From the foregoing observations, we 

 may conclude, that the air thermometer 

 requires no correction of its indications ; 

 that the accuracy of the mercurial 

 thermometer is not materially affected 

 by the inequalities of the expansions 

 of the mercury, in ordinary ranges of 

 temperature ; that the expansion of al- 

 cohol is pretty uniform, until about 30 

 R. or 100 F. : above that point its ex- 

 pansions become more irregular ; but it 

 has the advantage over every other liquid, 

 of marking the lowest degrees of natural 

 or artificial cold hitherto observed. 



The irregularities affecting pyrome- 

 ters, except from alteration in the size of 

 the substances supporting the bars, are 

 extremely minute. The experiments of 

 De Luc and Roy would lead to the con- 

 clusion, that the expansibility of solids 

 is not quite equable. Roy thinks that 

 this slight irregularity may be apparent 

 rather than real; but Lavoisier and 

 Laplace state, that the expansions of 

 solids keep pace with those of the mer- 

 curial thermometer, from the freezing 

 to the boiling point of water ; and Petit 

 and Dulong assert, that the expansion 

 of metals is progressive above 212 F. 



These irregularities, if they exist, are 

 so minute, as not, in any ordinary prac- 

 tical purpose, to affect the indications 

 of pyrometers; and at high tempera, 

 tures extreme accuracy is seldom of 

 much consequence. 



The differences arising from two py- 

 rometers of different materials may be 

 corrected by Table IV. in the appendix 



Phil. Trans, vol. Ixvii. 



t Manchester Memoirs, vol. v. p. 599. 



J Edin. Phil. Trans. Ann. de Chim. t. xliii. 



|| Annales de Chim, et Phys, t. ii. 



