404 Scientific Proceedings^ Royal Dublin Society. 



By the method of heating adopted, tliere is no difficulty in making the 

 temperatures in tlie second series agree with those in the first with an error 

 of less than 0-05°, and the difference is in most cases much smaller. 



2. By the other method the volumes of a gram of liquid can be determined 

 with a single tube ; but the experimental difficulties are considerably greater, 

 and it is doubtful whether quite the same degree of accuracy can be attained. 

 It will be sufficient to state here that, at each temperature, four readings of 

 tlie volumes of liquid are taken with the wliole of the vapour and different 

 and increasing measured lengths of the column of liquid heated to the high 

 temperature, the lower part of the liquid being maintained at a constant low 

 temperature by a current of water. For the last reading, the whole of the 

 liquid as well as the vapour is heated to the liigh temperature. A full 

 description of tlie modified apparatus and of the method of calculation is 

 given in the original paper.^ 



Specific Volumes of Saturated Vapour. 



The specific volumes of saturated vapour were determined either by direct 

 or indirect methods, or by both. 



Indirect Method. 



The specific volumes of unsaturated vapour were determined at a series 

 of temperatures, a large number of observations under different pressures 

 being made at each temperature. Plotting the volumes against the pressures, 

 the volumes of saturated vapour are given by the intersection of the 

 isotliermal curves with the straight lines of vapour-pressure. For these 

 determinations two forms of apparatus were employed. 



1. For lyre&mres lower or not much higher than one atmosphere, the 

 Gay-Lussac-Hofmann vapour-density apparatus was modified in such a way 

 that the external pressure could be altered, and a series of observations made 

 at each temperature with the same quantity of substance.^ 



The apparatus in its final form is shown in fig. 9. In place of the 

 ordinary shallow mercury trough, a cylindrical gas-jar A is employed. This 

 is closed by an india-rubber cork perforated with one large central hole, 

 widened near the top, and two small holes near the sides. The vapour- 

 density tube, B, passes through the central hole ; and its lower end is 

 immersed in the mercury in A. Through the small holes in the india-rubber 



' Trans. Chem. Soc, lix., p. 37, 1891. 



" Eamsay and Young, Trans. Roy. Soc, clxxviii., p. 57, 1887. Young suid Thomas, Proc. 

 Phys. Soc, xiii., p. 658, 1895. 



