Flotation of Minerals, Part II. 319 



These two experiments show that attachments of carbon 

 dioxide still cling to the particles of sulphide on settling be- 

 neath the surface of the liquid, and after the visible gas bubbles 

 forming the scum have become disengaged. The analysis or the 

 disengaged gas in the previous tests show that it is mainly 

 composed of nitrogen. Assuming that a proportion of oxygen 

 that would give with the nitrogen found a mixture correspond- 

 ing to atmospheric air, was originally present, the fact that the 

 gas evolved from the sulphides themselves when subjected to 

 reduced pressure is carbon dioxide, shows that the latter is 

 attached more persistently than is the air. 



7. An attempt was then made to collect any gas that might be 

 evolved from dry minerals on subjecting them to a low pres- 

 sure, in the following manner : — The apparatus was an auto- 

 matic Toepler pump, designed to collect the gas from the 

 apparatus being exhausted.! The mineral to be tested was en- 

 closed in a glass flask, and connected with the pump. The 

 pump was then worked until no more gas could be obtained at 

 ordinary temperature, and the last portions of the gas were col- 

 lected for analysis. The flask containing the mineral was then 

 heated by immersing in a water bath, containing boiling water, 

 the pump being kept continuously working, and the further 

 quantities of gas obtained were collected and analysed. 



At first the pump was used with a desiccating apparatus 

 inserted between the flask containing the mineral and the 

 pump, but as the desiccating agent used (phosphorus pentoxide 

 was first tried, and then dehydrated copper sulphate) in each 

 case apparently adsorbed and retained to a greater or lesser 

 extent portion of the gas first evolved, the pump was finally 

 used without the desiccating apparatus, the minerals being 

 previously dried as far as as possible by keeping them in a 

 desiccator for several days. The vacuum obtained under these 

 conditions was, however, not so complete as when the desiccat- 

 ing apparatus was used. 



Table A shows the results of these tests. Column III. of this 

 table shows the composition of the last portions of gas 



I "An Automatic Toepler Pump designed to collect the gas from the apparatus being 

 exhausted." Bertram D. Steele, D.Sc, Proc. of Physical Society, London. Vol. xxii. ; also 

 Phil. Mag.. June, 1910. 



