442 DR LYON FLAYFAIR AND MR J. A. WANKLYN ON A MODE OF 
after noting the barometer, is sealed with the blowpipe flame. On cooling, it is 
weighed, the temperature of the balance case and the barometric pressure being 
noted. The difference between the two weighings, added to the weight of the air 
displaced by the sealed flask, gives the weight of the vapour in the flask when it 
was sealed. The volume of the vapour at the time of sealing was equal to the 
capacity of the flask, its temperature was that of the bath, and the pressure to 
which it was subject was that indicated by the barometer. It only remains to 
ascertain the capacity of the flask, in order to compare equal volumes of air and of 
the vapour. For this purpose the flask is filled with water or mercury, and the 
capacity determined by weight or measurement of these liquids. ‘The weight of 
the volume of air can now easily be determined by reference to tables constructed 
for this purpose. After applying the corrections previously described under Gay- 
Lussac’s process, we can readily fix the vapour density of the liquid, as we have 
all the elements for the ratio of the weights of equal volumes of the air and 
vapour at the same temperature and pressure. Both the processes of Gay- 
Lussac and Dumas require for their successful execution, that the vapour 
densities should be taken at 30° to 40° C. above the boiling point of the body 
operated upon. This restriction of their application at once excludes the numer- 
ous class of substances which will not bear this elevation of temperature without 
decomposition. Before we describe the methods which we employ to determine 
the vapour densities of liquids at temperatures below their boiling points, it is 
necessary to discuss the difficulties inherent in the processes already described, 
and which compel the operation to be performed at the highest practicable tem- 
peratures. 
Vapours do not expand equally, for equal increments of heat at temperatures 
close to their point of condensation, and even at temperatures considerably 
elevated above this. A vapour only a few degrees above its temperature of con- 
densation, has a tendency to condense on the surface of solids plunged into it, 
and on the sides of the vessel containing it. Even among the permanent gases, 
such as carbonic acid, protoxide of nitrogen, and sulphurous acid, there is a higher 
coefficient of expansion between 0° and 100° C. than there is for air, although at 
these temperatures, and at ordinary pressures, they are still several atmospheres 
removed from condensation (REGNAULT).* At the same time, be it observed, 
that these gases are among those of the permanent gases, which are most easily 
liquefied; and this fact has led to the belief, that vapours also have their 
coefficient of expansion abnormally high, as they approach temperatures at which 
they pass into the liquid state. In fact, it is a matter of common observation, 
that the vapour density of a body taken near the boiling point is frequently 
higher than it should be by theory. 
* Reevautt, Annales de Chim. et Phys. (1842), 3™° séries, v. p. 80. 


