366 Mr. J. Rose-Tnnes and Dr. S. Young on the 



o 



volumes; and a line running through the points might 

 apparently end at zero difference. But though this result 

 might be accepted as consistent with the experimental 

 evidence if there were independent grounds for believing in 

 it, it cannot be considered as the most probable judging 

 solely from the diagram ; we should be rather led to believe 



that even at infinite volumes the value of — : . for isopentane 



remained larger than that for normal pentane. 



The above results respecting a and t are chiefly interesting 

 because they seem capable of throwing some light on the 

 vexed question of the influence exerted by difference of 

 chemical structure on the thermal properties of a substance. 

 Concerning this matter very little is known at present ; but 

 it is common knowledge among organic chemists that two 

 substances may have the same chemical composition and 

 show practically the same behaviour whilst in the condition 

 of rare vapour, and yet they may differ considerably as to 

 their thermal properties in the liquid state. The great field 

 of observation in which the substances lie between the con- 

 ditions of a rare vapour and a common liquid has been left 

 almost entirely unexplored. This gap in our knowledge 

 makes it impossible to say in what precise manner the differ- 

 ence between two isomeric substances originates ; whether it 

 arises conjointly with the first deviations from Boyle's law, 

 or whether the difference remains inappreciable even with 

 increasing density until we reach the neighbourhood of the 

 critical point. We may put the problem more precisely as 

 follows : — If we imagine the pressure given by a series of 

 ascending powers of the density, what is the lowest power of 

 the density which has different coefficients for two isomeric 

 substances ? 



We are now able to answer this question with a fair amount 

 of exactness in the case of the two isomers, normal pentane and 

 isopentane. If, as seems most probable, there is a difference 



between the — ^ for normal pentane and the — T for iso- 



pentane, even at infinitely large volumes, this shows that the 

 coefficients of the second power of the density in the expan- 

 sion of p must be different for the two substances. On the 



other hand, if there is no difference between the — « for 



-j avr 



normal pentane and the — ^ for isopentane at infinitely large 



volumes, then the coefficients of the second power of the 

 density in the expansion for p must be the same, since t has 



