23S 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



[AUGOST, 



But I must, in a very especial manner, testify my fjratitude to 

 niv tVienil M. Iznrn, fur the zeal and complete devotion with which 

 111' lias .lidc'd me in tliis lonsj series of laliimrs, some of which were 

 not without ilanu'er, and all were very trouhlesome, as well as m 

 t!ie loMij and tedious numerical calculations which were tlie con- 

 se(pieu<-e of them. Hy the aid of his active co-operation, I have 

 lit'eii ahle to terminate these labours in much less time than it 

 woiilit liave been pnssil)le for me to have done it if I had been re- 

 duced to my own personal efforts alone. 



First Memoir. — on the dilatation of elastic fluids. 

 Part I. — Dilatation of Atnwipheric. Air, under the ordinary 



Pressure of the Almnsjihere. 

 M. Rejjnault commences his memoir, by remarkinsj that there 

 is, in ])liysical science, no numerical element wliicli has been sub- 

 mitted to a {i^reater number of experimental determinations than 

 tlie co-efficient of dilatation of atmos])lieric air, and that never- 

 tlieless we cannot yet say that this co-efficient is Ivnown to us with 

 sufficient precision. The experiments of tlie elder physical philo- 

 sophers gave numliers so different from each other tliat no use 

 VAW be made of tliem. The greater part of the circumstances 

 wliicli influenced the phenomenon were unknown to them. 



Tlie ex])eriments of jNI. Gay Lussac (Annales de Chimie, 1st 

 Series, tom. xliii., p. 137. Biot. Traitc de rhysique, torn, i., 

 p. 182), seemed to have settled the <|uestion finally. He showed 

 by a great number of experiments that between 0° and 100^ (32° to 

 212' Fahrenheit) the co-efficient of dilatation was the same for 

 all gases, and for vapours, when they were at some distance from 

 their point of condensation, and that its value was 0'375."' 



This co-efficient was adopted liy all physical philosopliers, and 

 employed in calculations, until in these latter years a Swedish 

 philoso])her, M. Rudberg, cast a doubt upon its exactness. By a 

 series of experiments made with care, j\I. Rudberg endeavoured to 

 show that the co-efficient of M. Gay Lussac was much too large, 

 and that its true value was comprehended between 0'364. and 

 0-3(i.>. 



The experiments of M. Rudberg are then described at length, 

 by M. Regnault. These experiments were originally published in 

 two memoirs contained in Poggendorif 's Annals, vols. xli. and xliv., 

 and the English reader will find them in the valuable Scientific 

 Memoirs, edited by Richard Taylor, vol. i., pages 507 and 514.. 



Rudlierg terminates his second memoir by an im])ortant remark, 

 which had already been made in 1803, by Gilliert (Gilbert's 

 Annals, vol. xiv., page 267), but had been entirely forgotten, viz : 

 that the experiments of Messrs. Dalton and Gay Lussac, which 

 had been regarded as having given almost identical results, diifered, 

 on the contrary, very much. In fact, in the memoir of Ualton 

 (Memoirs Soc. Manchester, 1st Ser., Vol. v., Part 2, p. 598), he 

 says : — 



" I have repeatedly found that 1000 parts of common air, of 

 the temperature i-i°, and common pressure, expand to 1,321 parts 

 of the thermometer ; to which, adding four parts for the corre- 

 sponding expansion of glass, we have 325 parts increase upon 

 1000 from ■Vj'^ to 212°, or from 157 of the thermometer scale 

 ( FiilirniheitJ." It is evident that the volume of air here assumed 

 as tlie unit, is that of air at 55° Fahrenheit, or 12-78 cent. If, on 

 tlie contrary, we take for unit the volume of air at 0° (32° Fah- 

 renheit), and put the dilatation between 0° and 100° =: lOOo, the 

 results of Dalton give 



1 -I- 12-78a: 1 -\- lOOo : : 1000 : 1325; whence 100a = 0-392. 

 This, then, is Dalton's true result. In truth, Dalton himself 

 does not appear to have observed the error which had slipped into 

 his calculations, for he says in his new system of chemical phi- 

 losophy : — " The volume of air, according to the experiments of 

 M. Gay Lussac and Mine, being 1000 at 32' Fahrenheit, becomes 

 1376 at 212° Fahrenheit." 



In a note JNI. Regnault notices a series of experiments upon the 

 same subject, made about the same time with his own, by Pro- 

 fessor Magnus of Berlin. An extract from Professor ^IagIlus' 

 memoir will be found in the Annales de (Chimie et de Physique, 

 3rd Ser. tom. iv., page 330 ; and a second memoir upon the same 

 subject, tom. vi., page 353. 



IVI. Regnault then proceeds to give his own method of experi- 

 menting, and the details of his experiments. 



These methods were five in number. In the first four, the dila- 

 tation of the air was deduced from the observed changes in its 

 elastic force at the temperatures of 0° and 100° cent., assuming as 



* The results arrived at by Mr. Dalton, about the aame time (Memoirs Lit. and Phil. 

 Soc. of Manchester, Ist. Ser., Vol. v., part '2, p. .'il)8), appeared to give a co-efficient, iden- 

 tical, or nearly so, with that of Gay Lussac, {U-.i726), and coiihrmed his assertion as 

 to the equal dilatation of dilTererit gases, so tliat Mr. Dalton himselt adopted the co.«fficient 

 found by M. Gay Lussac. 



true the law of Mariotte, that the elastic force of a gas varies 

 inversely as its v(dume, when the temperature remains the same. 

 The fifth method was an attem]it to measure directly the augmen- 

 tation of volume due to the cliange of temperature. 



Thefirxt methiid was similar to that used by Rudberg, in his first 

 series of experiments, and by Dulong and Petit, in their com- 

 parison of mercurial and air thermtmieters. 



The apparatus consisted of a glass cylindrical reservoir, from 

 25 to 30 millimetres in diameter, and about 110 millimetres long, 

 containing from SOO to 1,000 grammes of mercury. To this was 

 soldered a capillary stem, of which the diameter varied in the 

 different experiments, from i to 2 millimetres. This was bent at 

 right angles, at stone distance above the reservoir, and drawn out 

 to a fine point. The reservoir and the greater ])art of the stem 

 were immersed in a vessel of water boiling under the usual atmo- 

 spheric pressure, and filled with perfectly dry air, by exhausting it 

 from 25 to 30 times, by means of a small pump, and re-filling it 

 each time with air which had passed through two tubes, eacli one 

 metre in length, filled with pumice-stone, saturated with concen- 

 trated sulphuric acid. This being done, the apparatus was suffered 

 to stand from half an hour to an hour, the water being maintained 

 in full ebullition ; the end of the capillary stem was then closed by 

 the blow-jnpe, and the height of the barometer noted. The 

 reservoir, with its stem, was then inverted upon a stand, so that 

 the point of the stem dijiped to some distance in a cup of mercury, 

 the cup was broken off under the mercury, and the reservoir sur- 

 rounded with iiounded ice, and left in its condition for an lioin or 

 more, until the wjiole of the air (now contracted so as to fill only 

 a portion of the reservoir) was reduced to the temperature 0°. 

 The end of the stem was then again closed by a little wax, the 

 barometer again noted, the position of the surface of the mercury 

 in the cup marked by a point adjusted by a screw — the cup re- 

 moved, and the reservoir and its contents suffered to take the 

 temperature of tlie surrounding air. The height of the mercurial 

 column above the level of the mercury in the cup was then 

 measured by the cathetometer. The reservoir and its contents 

 were then weighed, entirely filled with mercury, first boiled to free 

 it from air and moisture, the point again immersed in mercury, 

 and the reservoir surrounded with ice. At the end of one or two 

 hours, when it was satisfactorily ascertained that the whole appa- 

 ratus had taken the temperature 0°, the ice was removed, the mer- 

 cury which was discharged by the rise of temperature was 

 received in a cajisule, and the apparatus placed in a boiler, as at 

 first, and brought to 100°. The mercury expelled was collected in 

 the capsule, and the height of the barometer at the moment of 

 ebullition noted. By tliis means, all the data necessary to calcu- 

 late both the dilatation of the air, and that of the glass vessel 

 which contained it, were given. 



In performing these experiments, M. Regnault observed a 

 serious cause of error. When the point of the stem was broken 

 under the mercury, he observed that a small quantity of air 

 leaked into the reservoir, even when the point was plunged to the 

 depth of -if metre under the mercury. This air was a ]uirtion of 

 that which remained in contact with the glass tube, which not 

 being wetted by the mercury, allowed, as it were, a tube of air 

 from the point to the surface. This difficulty was obviated by at- 

 taching to the glass stem, plates of well-cleaned brass, to which 

 the mercury adhered, and tlius the entrance of air was prevented. 

 In addition to this, a layer of sulphuric acid was sometimes poured 

 upon the surface of the mercury, before the point was broken, and 

 was carefully removed before the point was again closed. Equal 

 care was taken to in-event the air enfilming the pincers used to 

 break the point, from getting access to the interior. In this 

 method fourteen experiments were tried, the mean of which gives 

 for the volume of 1,000 measures of air, at the temperature of 0°, 

 when heated to 100°, 1-36623. 



The highest number obtained in any experiment, was 1-36689 

 The lowest ... ... ... ... ... 1-36519 



The difference is 0-00140 



or about Tmt of the mean. 



The lowest number was above the mean result obtained by Rud- 

 berg. M. Regnault believes that this may probably be due to the 

 phenomenon of the entrance of the air upon breaking the point 

 having taken place in the experiments of the Swedish professor, 

 and he remarks that the error would be greater in proportion as 

 the quantity of air operated on was less. He also states that he 

 believes the first experiments of his own series were affected by 

 this phenomenon, and as an evidence of this states, that from the 

 moment that he succeeded in preventing it entirely, no experi- 

 ment gave a number below 1-3659. 



