364 Mr. J. Kose-Innes and Dr. S. Young on the 



An examination of this table shows that r remains fairly 

 constant for all large volumes down to about vol. 8. The 

 actual numbers obtained vary a good deal ; but these variations 

 are sometimes in one direction and sometimes in another, and 

 there is no steady increase or decrease. It appears, then, 

 that all the values of r above vol. 8 could be treated as the 

 same without introducing any serious error ; this occurred 

 likewise in the case of isopentane. What is still more note- 

 worthy is that the same constant value of r could be used for 

 both normal pentane and isopentane, keeping within the 

 limits of experimental error. The mean value of t for all 

 volumes above 8 was found to be 842*4 for isopentane ; it is 

 838*5 for normal pentane ; and the intermediate value 840 

 could be used in both cases without introducing any error 

 greater than the unavoidable errors of experiment. 



When we pass on to the neighbourhood of the critical 

 point, the value of r diminishes steadily as the volume 

 decreases. For the critical volume itself t is about 807, and 

 for vol. 2 it has sunk to 488. 



The most important conclusion arrived at in the case of 

 isopentane was that the molecular pressure a does not follow 

 a continuous law, but passes abruptly from one law to 

 another somewhere about vol. 3*4 (Phil. Mag. xliv. p. 79). 

 This inference was based on the study of a diagram in which 



the quantity — ^ was plotted against t>~*, and there appeared 



to be considerable evidence of discontinuity in the neighbour- 

 hood of the volume already mentioned. Of course it is 

 impossible to prove discontinuity of slope by means of a 

 series of isolated points, but it is suggested very strongly ; 

 and even if there be not discontinuity in the true mathematical 

 sense of the term, there seems to be such a rapid change of 

 behaviour as to amount practically to the same thing. 



It was therefore a matter of some interest to discover 



whether the diagram obtained by plottino- — - against v~i in 



the case of normal pentane would exhibit the same peculiarity. 

 The diagram is given on p. 363, and it is easily seen that we 

 have here a similar suggestion of discontinuity in the slope of 



— 2* this occurs somewhere about vol. 3*4, as with isopentane. 



In attempting to find a formula for the pressure of normal 

 pentane we are therefore confronted with the possibility that 

 we may require two distinct algebraic equations. We may 

 simplify the problem considerably by confining our attention 



