Notices respecting New Books, 53 



impossible, on account of the necessity of stationary elevated tem- 

 peratures. However, experiments of this kind might be made with 

 success up to the temperature of boiling water, determining by means 

 of the balance the weight of gas which fills the balloon under differ- 

 ent pressures, and at the fixed temperatures of melting ice and boil- 

 ing water. M. Regnault describes the apparatus, composed essen- 

 tially of two copper globes, which he destines to experiments of this 

 kind ; but the experiments themselves have not yet been executed. 

 M. Regnault attaches the more importance to them, that they not 

 only allow of the determination of the influence of that important 

 element, the temperature, but that they may moreover serve to con- 

 trol the method followed in the other experiments, to which the 

 objection might be made, that the anomalies perceived in atmosphe- 

 ric air are due, at least in part, to a condensing action of the surface 

 of the glass tube which contains the air. 



To sum up the foregoing, we see that the very complete and pre- 

 cise researches of M. Regnault conduct to the conclusion, that neither 

 the law of Gay-Lussac on the dilatation of elastic fluids by heat, 

 nor that of Mariotte on the constant relation which exists between 

 their elastic force and their volume, are laws of nature, but that 

 they are only approximative, and consequently only true between 

 certain limits. The aberrations manifest themselves for both the laws 

 in tolerably similar circumstances, and demonstrate the intimate rela- 

 tion which exists between the two classes of phsenomena. Thus 

 the coefficients of dilatation are greater in proportion as the density 

 of the gas is more considerable ; the same holds of the compressi- 

 bility* hydrogen being the sole exception in both cases. Carbonic 

 acid gas dilates more, and likewise experiences a greater compression, 

 than atmospheric air in the same circumstances. Many other paral- 

 lels of the same kind might be made, but I prefer confining myself, 

 in concluding this account, to dwelling for a moment on the conse- 

 quences which, it appears to me, may be deduced from M. Regnault's 

 researches, in reference to the constitution itself of elastic fluids. 



It has long been said that in gaseous substances the influence of 

 their special nature is no longer perceptible, and that their physical 

 properties, excepting the density, depend alone on an agent common 

 to all, namely heat, and that it is for this reason that the laws they 

 present are general and the same for all, such, in particular, as Gay- 

 Lussac's and Mariotte's laws. Error in the consequences, therefore 

 error in the point of departure. 



But what M. Regnault's labours appear to prove, is, that the less 

 the density of the gases, the more general and similar appear the 

 laws they present. Only it is necessary, in the case of those of 

 which the low density does not result, as in hydrogen, from their 

 special nature, that an elevation of temperature should compensate, 

 as far as possible, for the loss of elastic force to which the diminu- 

 tion of density subjects them. In one word, what is required is, that 

 the atoms should be separated as widely as possible, and the elastic 



