( 183 ) 



Chemistry. — ''On the cowsd of tlw, I'T-Unea for constant cnn- 

 centration in the systnn ether -anthra(jumone." By Prof. A. Smits 

 and J. P. Tkeub. (Coimuunicated by Prof. J. D. van der Waals). 



Already in 1903 ^) Van der Waat.s indicated by means of PT- 

 sections for constant concentrations how in the iieighbonrliood of 

 the critical end-points 7^ and q (he meeting of the lines for WiV/-//«/(/, 

 and liquid-vapour wonld have lo take phice. 



Theory, however, was in liiis respect far in advance of experiment, 

 so that a perfectly nnexpecled |)ecidiarity found afterwards hv one 

 of us (S.), in consequence of which the said lines must partly have 

 another position with respect to each other than van dkr Waals 

 had supposed, could not be taken into account in these considerations, 

 and for this reason it was very desirable to investigate this point 

 more closely experimentally. 



The just-mentioned peculiarity consists in this that for some mix- 

 tures lying on the ether-side of the critical end-point p, far above 

 the critical temperature, ^vhich was of course observed here for 

 unsaturate solutions, three-phase equilibrium (/9-f-^-]-(^') appeared again. 



This behaviour points to the fact that the line solid-fluid for a 

 concentration on the left of the first critical end-point does not meet 

 the border-line liquid-vapour tnnce, as van der Waals supj)Osed, but 

 four times, so that the Pl'-iigiwe for this concentration becomes 

 about as indicated in fig. 2, which is of course preceded by contact 

 on the right, as drawn in tig. 1, which contact of course takes place 

 for the vapour most rich in ether of the second part of the three- 

 phase region. 



Since in the case that the critical phenomenon is observed for 

 an unsaturate solution, the plaitpoint K lies between the liquid point 

 L and the vapour point G, we see that the peculiar phenomenon 

 discussed here is iji connection with the enormously large difference 

 which must exist for this system between the plaitpoint temperature 

 and the maximum temperature. 



As one of us (S.) has set forth in his first communication ") on 

 the (/-'y'V-lines for solid-fluid by means of the formula: 



T 



dp 



dT 



sf/x 



1) These Proc. Oct. 1903, p. 230; Nov. 1903, p. 357, 



2) These Proc. May 1906, p. 9. 



