Thermal Properties of Normal Pentane. 367 



already been shown to have the same value for the two sub- 

 stances at infinite volumes ; and the lowest power of the 

 density which has different coefficients for the two isomers 

 must be the third. 



It w r as thought advisable to test these conclusions by a 

 different method. In a former paper by one of the authors 

 (Phil. Mag. xliv. p. 80 ; see also Phil. Mag. xlv. p. 105) it 

 was shown that in the case of isopentane we might reproduce 

 the original observations very closely by putting 



* v \ v + k — gv~ 2 ) 



gv~ 2 ) v(v + k) 9 



where R, <?, k, g, and I are constants. If we assume that this 

 formula holds also for normal pentane, and if it be true that 

 the difference of pressure between normal pentane and iso- 

 pentane at the same temperature and volume varies ultimately 

 as the third powder of the density, then we should be able to 

 reproduce the experimental data for normal pentane by 

 means of the above formula, giving to R, e, and I the values 

 already found for isopentane. We may accordingly take 

 11 = 863-56, £ = 7-473, 1= 5420800, and we still have the 

 constants k srndg left at our disposal to meet the requirements 

 of the normal pentane data. On examining the observations 

 given in Series I. of this paper we find that we can con- 

 veniently put k — 3*135, #=6'695, and we have to test how 

 far the formula with these constants reproduces the experi- 

 mental results given in Series II., III., and IV. In order to 

 institute an effective comparison between theory and observa- 

 tion a diagram was made in which pv was plotted against v~i ; 

 the calculated isothermals were drawn as continuous lines, 

 while the experimental values were put in as dots. An 

 examination of the diagram shows that a fair concordance 

 between calculation and experiment has been secured ; but 

 the agreement is not so good as could be wished. Deviations 

 amounting to 1 per cent, are not uncommon, and in places 

 they approach 2 per cent. If we have regard to the differ- 

 ences which often occur in inquiries of this kind between the 

 results of independent observers, we might conclude that the 

 above deviations are unimportant, and that R, e, and I were 

 really the same for the two pentanes as supposed. But it 

 seems more likely that the deviations are too large to be 

 neglected ; hence, the most probable inference is that the I 

 for normal pentane is not the same as the I for isopentane, 

 thereby confirming our former conclusion. 



2 C2 



