4 



Drs. W. Ramsay and S. Young. 



[Jan. 6, 



similar investigation on ether. We have also finished a study of 

 the thermal properties of methyl alcohol. 



In consequence of a recent publication by Wroblewski, of which 

 we have seen only the abstract (' Deiitsch. Chem. Gres. Berichte,' 1886 

 (Referate), p. 728), we deem it advisable to communicate a short 

 notice of an investigation in which we are at present engaged. 



We find that with the above-mentioned substances, acetic acid 

 excepted, whether they are in the liquid or gaseous state, provided 

 volume be kept constant, a simple relation holds between pressure 

 and temperature. It is 



p = bT-a. 



This is evidently a simple modification of Boyle's and Gay-Lussac's 

 laws; for at low pressures, where volume is large, the term a 

 approaches and finally equals zero, while b diminishes and finally 

 becomes equal to the value of c, calculated from the ordinary 

 equation, . 



= °5 

 ^ v ' 



We have as yet only had time to apply this formula with ethyl 

 ether to the liquid state ; and as we are not yet quite certain whether 

 the relation holds when 1 gram of ether occupies volumes between 

 4 and 20 c.c, we are at present engaged in measurements of volumes 

 and pressures at temperatures between 220° and 280°. Assuming the 

 above relation to be true (and it is at all events a close approximation 

 to truth), it is possible to calculate those portions of isothermals 

 included within the liquid-gas area, and represented in Andrews' 

 diagram by horizontal straight lines. We have calculated a few of 

 these isothermals for ether, and find that the areas above and below 

 the horizontal lines (see woodcut) are equal, when measured by a 

 planimeter. 



Reserving a full discussion of the subject until the completion of 

 our experiments, we would here point out the similarity between 

 the equation p = bT — a, and those proposed by Clausius and by van 

 der Waals to represent these relations. Clausius's formula is 



RT c 

 and van der Waals' m — ^ ^ 



1 v — I) V" 



In these formula? Clausius's <x and c are equivalent to van der 

 Waals' b and a respectively, but R has a different signification. 



We find that a somewhat similar formula agrees better with expe- 

 riment than either of the above ; it is 



