Relations of Liquids and their Saturated Vapours. 99 



all liquids which belong to the ether-node. I find that by 



taking C from Professor Andrews's observations on latent heat 



at the boiling-point, which corresponds nearly to the r tempe- 

 rs 



rature, and computing therewith -7-, a value of about 21 always 



ft 



appears in those cases that are likely to belong to the ether-nede 



class. 



§ 47. A higher value seems to belong to liquids that belong 



to the water-node; thus ~- for water is 30*4, and for pyroxylic 



spirit 29*25. For alcohol a still higher value is found, viz. 35*7; 

 and its line cuts the zero-line z q (fig. 1) still further down than 

 the water-node. 



§ 48. The latent heat of chloroform, of sulphuric ether, and of 

 some other bodies have been determined by Regnault up to high 

 temperatures, so that we are enabled to test the presumed law 

 &« a higher value of t. I have computed the values of C cor- 

 responding to all the values of X (making use of the empirical 

 formula for Q), and laid them off as ordinates to the respective 

 temperatures ; then, having drawn the curve by flexible straight 

 edge, the value of C at the respective t temperatures is read off 



and -j- computed. 



Let 120° = the new t temperature of sulphuric ether ; a 

 parallel to the axis drawn through this point (Plate III.) cuts 

 the chloroform line in 157°. At these temperatures the 

 respective values of C are 73*2 and 49 '4, which give 17*23 



and 17*31 as the respective values of -j—, an accordance that is 



surely very remarkable. 



§ 49. Let us now compare C at different temperatures of the 

 same liquid with the volume v at those temperatures. First, 

 sulphuric ether, the expansion constants of which are determined 

 from Pierre to be in conformity with the general law (7 = 224, 

 P o =3*01). I take from the curve of C 91*7 at 20°, and 78*0 

 at 100°, — the volumes at these temperatures being respectively 

 1-031 and 1-199. Hence we have 13*7, the difference of C, 

 equivalent to 0*168, the corresponding difference of v. In the 

 same proportion, 13*7 : 0*168 = 78*0 : 0*957, the augmentation 



from 100° up to the maximum volume at transition, if — k — is 



ov 



constant. This gives 2*156 as that maximum when C is reduced 

 to zero. Now the temperature corresponding to this volume, 

 computed from the above value of y and P, is 217°. The ob- 

 served temperature of transition was 221° (§ 29), and maximum 



H2 



