398 



CATALOGUE, HEAT, CHANGES OF FORM. 



alcohol. Hence he obseiTcs, that the logarithm of the 

 elasticity should vary as the temperature. Encycl. Br. Art. 

 Steam Engine. He could, however, discover no sensible 

 elasticity in ilcohol at 32° ; nor could Betancourt Molina. 

 Dalton pursues Robison's idea of the logarithmic, with 

 some alterations: he made experiment? both under the 

 air pump and with the Torricellian column ; he found that 

 a difference of ll^" increased the elasticity 1.4925 times 

 at 3J°, and 1.2425 times at 212", hence he infers that 

 .015 is to be deducted from the ratio for every such inter- 

 val, and continues his table both ways. But it is certain 

 that this cannot bs the law of nature, since about 394° th« 

 elasticity would become uniform, and then decrease, if the 

 law were true. He says that Betancourt and Robison 

 make the elasticity too great in high temperatures from 

 the e.\trication of air: but the fact is, that when the 

 greatest care has been taVen to avoid it, the elasticity has 

 appeared nearly the same, and the circumstance, if it had 

 taken place, would have been very immaterial. Indeed, 

 the only support of Dalton's measures above the boiling 

 jx«nt is the law, which he has imagined for the expansion 

 of other vapours ; he tays, that their" elasticity is always 

 equal to that of steam, at a given difFerence of temper- 

 ature above or below them : and some experiments, that 

 he adduces, agree exactly with the law j but it is utterly 

 incredible that an expansive force of 7 tenths of an inch 

 vihich thevapourof alcohol ought to have at the freezing 

 point, should have entirely escaped both Betancourt and 

 Professor Robison. Still, however, his rule for the force 

 of different vapours must be allowed to be a very valuable 

 approximation at temperatures between 50° and 220°. 



A much simpler formula will agree extremely well with 

 all Dalton's experiments on water, and with the mean of 

 all the best experiments that have been made by others in 

 higher temperatures. It is this, the elasticity of steam in 

 atmospheres of 30 inches of mercury is rf:r:(l-f-.002(}/")', 



y being the degrees of Fahrenheit above a 1 -a", whence we 

 .1 



have /iz ' for the elevation of the boiling point with 



.0029 



an increase of pressure. If we reckon f from 32", we 



shall have the elasticity in inches of mercury nearly .1781 



(i-f-.ooe/)' ; and for the ejevation or depression of the 



boiling point, if e be the elevation of the barometer 



above 30 in inches, we shall have for small variations /zi 



* e 



»^ r: ).642e. Deluc makes ih* cor- 



7 X .10 X -oo'^g .009 

 rection 1.59c, Shuckburgh l.7oe, the mean is i.64 5e, 

 which agrees very singularly with the calculation. Ac- 

 cording to Dalton's principles the formula may be ac- 

 commodated to any other vajiour, by reckoning / from 

 some other constant point of the scale ; as —5° for 



alcohol, 50° for muriate of lime. 



I.4113-f.005r 

 Schmidt's formula is c-=.r in hundredths 



of a French inch of mercury, r being the temperature 



in degree* of Reaumur. This is nearly equivalent to 



1.163-f.002l/ 

 / in hundredths of an English inch,/ being 



the degrees of Fahrenheit reckoned from 32° ; or to 



1.33S8-I-.004C 



c for the degrees of the centigrade ther- 



mometer. 



Prony's formula for Bctancourt's experiments, is ri- 

 diculously complicated, and yet not at all accurate. 



Soldner gives, for expressing Dalton's numbers, the for- 

 (■6f)2— /). (212— /■) ,. 



mulaen:l. 30.13. ■—^ — .He accommodates 



52042 

 similar formulae to other scales, and deduces from them 

 others for the determination of the heat of boiling water 

 under different pressures. , 



This however is only an approximation to Dalton's 

 principle, which from the properties of the logarithmic 

 curve, leads to a formula of this kind, eiz.016l373 — 



(1.0365 — .00008/). I. (1.0365— .00008/) .4343 



(.00008/). 



I have also found several expressions which for particular 



purposes may possibly be of some use, although thejr arc 



all superseded in general by the formula first mentioned ; 



1.5 

 these are, reckoning / always from 32°, c— 003/ , 



.006788/ *0«7e8/ 



e=:io /, e=:io (/+.000329/*— .oooooooi 



(.0551/— .000019^\ 

 10 /e— numb. 1. .3 0103 -f- 



.01541/— .000017/'— .000 000 008/', in tenths; and for 



atmospherical temperatures ez:.2-}-.007/4-.00016/', 



which is deduced from Dalton's table, but may perhaps 



be improved by making e^.ls4-.007/4- .000l({/"*. 



Comtruction of Thermometers. 



Braun's comparison of the scales of thermo- 

 meters. IN. C. Petr. VIl. pi. 18. 



VV'entz on dividing tlieniiometers with un- 

 equal tubes. Act. Helv. III. 105. 



Report of the committee of the R. S. Ph. tr. 

 ■ 1777. 816. 

 The stem of a thermometer being 1 00° colder than the 



bulb, the mercury will be about ij° lower in 180°. It 



ought always to be of the same temperature. 



The bulb being immersed an inch under water, the 



boiling point is raised .08°, which is about half as much 



as the same pressure would occasion if exerted by the air. 



