THE BOOK OF THE LANTERN. 41 



disconnected from the wash bottle, the water in the latter 

 may rush back into the retort and cause a small steam 

 explosion. 



This has certainly never happened to me ; but I b^ve 

 heard of cases where such a thing has occu: 1 



although it would probably be unattended by any id 1 1 tous 

 results, it might lead to a great deal of mess and 

 trouble. 



The retort may be left, until it has become nearly cold, 

 or at any rate until it is cool enough to be handled. The 

 branch pipe should then be unscrewed, and the retort at 

 once washed out with water warm water by preference. 

 This should be thoroughly done, and many changes of water 

 should be used, until the last wash water comes away per- 

 fectly clean. If the retort is left with the residue of the gas 

 mixture in it, the metal inside is very quickly corroded, and 

 the vessel does not last half so long as it does if it be at 

 once carefully washed. The branch and the india-rubber 

 connecting tubes should be washed out also. 



I have mentioned that* the commercial chlorate of potash 

 is contaminated with a certain amount of free chlorine. This 

 soon renders itself evident if the operator places his nose 

 near the delivery tube from the wash bottle when the gas 

 is coming off; for chlorine gas has a suffocating odour. 

 This is not the only disadvantage which it has in gas 

 making for lantern purposes, for it so acts upon the india- 

 rubber bag and the attached brass work that it quickly 

 leads to deterioration. By placing in the wash bottle 

 water something which will seize hold of this chlorine and 

 detain it, we shall avoid this last difficulty, and the 



