510 REPORT 1888. 



maintained that his results were rigidly accurate ; but as they are only 

 approximate the objection falls to the ground. 



Again, they give a table for the dilatations of water and other liquids 

 which Hirn had studied at a pressure of 11'5 metres of mercury ; and 

 deduce from these tables, by means of MendelejeflP's formula, the value 

 of k for water, from 100° to 200° ; for alcohol, from 50° to 200° ; for 

 ether, from 21-55° to 133'66° ; and for (essence of) terebenthene fx'om 40° 

 to 160°. 



The value of /.• for each substance in these determinations they find 

 not to be constant but to vary with t, increasing as t rises ; now the first 

 example chosen is that of water, which even below 100° Mendelejeff finds 

 to be a quite remarkable exception in being far removed from an ideal 

 liquid ; and with regard to the other examples (to which are added in 

 this paper of Bartoli and Stracciati, ammonia, nitrous oxide, carbon 

 dioxide, and ethyl chloride), it may be remarked that in Mendelejefi"s 

 paper, p. 131, tliere is this foot-note : — ' A subject for further experimental 

 investigation would be the solution of the question, in what manner and 

 how much the expansion of volatile liquids is changed at different 

 pressures. I should like to submit this theoretically important question 

 to an experimental solution.' 



TJie Continuity of the Gaseous and Liquid States. 



In nature we find no substance which is an ideal gas or an ideal 

 liquid ; however nearly a body may conform to either of these ideal states 

 within certain limits there are always deviations from such conformity 

 outside those limits ; so that volatilisable liquids, which in one direction 

 tend to obey MendelejeS''s law within certain limits of temperature, in 

 the other direction will, in general, under suitable conditions, tend to 

 conform with Gay-Lussac's law. There is, however, in all cases a more 

 or less considerable range of intermediate conditions in which the expan- 

 sion of the body is not even in approximate conformity with either law. 



Andrews ^ showed by experiments on various bodies that, when a con- 

 densible body is in the state of vapour, there exists a temperature — the 

 critical temperature — at and above which no amount of pressure could 

 make any appearance of discontinuity in the body acted on : no meniscus 

 could be observed indicative of the separation of the body — e.g., COq — ■ 

 into two distinct portions of different densities ; above the critical tem- 

 perature there is, as shown by graphic representation of the experimental 

 results, a continuous curve for each tempei'ature.^ Below the critical 

 temperature a suSicient pressure produces a meniscus, indicating a sepa- 

 ration of the body into a heavier and lighter portion ; and generally, at 

 temperatures below the critical, the diagram representing the variation of 

 volume with pressure showed a portion of the line as straight during 

 the compression so long as the less dense upper portion was visible ; after 

 this the curve shows a rapid increase of pressure with given decrease of 

 volume. The whole line of pressure-volume consists, roughly speaking, 

 at these lower temperatures of two curves joined by a straight line. 

 Above the critical temperature there is perfect continuity ; below it, dis- 

 continuity ; above, the relation between p and v (t constant) could con- 

 ceivably be expressed by an equation representing the continuous curve ; 



' Phil. Trans, ii. 1869, p. 575. ; and 1S76, p. 421. 

 ' Jlid. 1869, p. 575. 



