114 



Mr. G. Ansdell. 



[June 15, 



several important points to be considered. In the first place, it was 

 advisable to select those gases which could not only be easily prepared, 

 but whose physical constants had been thoroughly investigated ; it 

 would be an advantage to use gases having comparatively low critical 

 points, as the temperature would be much more easily kept constant, 

 and would consequently contribute to the accuracy of the results ; but 

 above all those gases should be chosen which would not be likely to 

 react upon each other in a liquid state, or at high temperatures and 

 pressures, for this would modify the results considerably, from the 

 probable formation of new compounds, &e. That this is likely to 

 occur is shown in Professor Dewar's experiments on the behaviour of 

 carbonic acid in presence of other bodies (" Proc. Roy. Soc," 1880), 

 where the carbonic acid often appeared to exist in a liquid state far 

 above its critical point. This was no doubt due to the formation of a 

 new compound ; at least, as it could not be pure carbonic acid, we can 

 only regard it as a compound of some kind formed under particular con- 

 ditions of temperature and pressure, and this supposition seems to be 

 confirmed by the experiment of carbonic acid in presence of camphor, 

 where the camphor undoubtedly formed a new body, for we know how 

 readily it combines with numerous substances such as hydrochloric 

 acid, &c, to form unstable compounds. For these reasons I chose car- 

 bonic and hydrochloric acid gases, as they could be easily prepared, and 

 their critical points had been very accurately determined, the former 

 by Andrews, the latter by myself, besides being bodies most unlikely 

 to be decomposed in each other's presence, more especially as they are 

 chemically saturated bodies, and therefore according to the new 

 chemical theory, most unlikely to form any addition or molecular com- 

 pound. 



The following method was adopted in the experiments, the 

 Cailletet pump being used as described in my former papers. The 

 carbonic acid was made by dropping pure strong sulphuric acid into a 

 saturated solution of potash bicarbonate, being afterwards washed 

 with distilled water, and dried by passing through four (J-^bes with 

 pounded glass and sulphuric acid. To check the readings of the air 

 manometer (which was the same used in my former experiments) and 

 also the purity of the gas, a tube was filled with the pure gas alone, 

 and the tensions at different temperatures and the critical point were 

 found to agree very well with Andrews' results. The hydrochloric 

 acid gas was prepared by the action of strong sulphuric acid on pure 

 chloride of ammonium, as described in my last paper (" Proc. Roy. 

 Soc," vol. 30), and was washed and dried with the usual precautions; 

 a tube was also filled with the pure gas to begin with, and the critical 

 point and tensions of the saturated vapour agreed as nearly as 

 possible with those obtained by myself two years ago. 



The purity of the individual gases and the accuracy of the air mano- 



