PREPARATION OF GASES 77 



difficulty; as with these proportions, if the gas is 'rushing,' 

 and the flame be lowered, the flow does not diminish for a 

 minute or two. Hence the probable behaviour must be anti- 

 cipated, as it were. This needs experience, and the beginner 

 should make his gas cautiously till he has gained it, which 

 will only take him rather more time to fill his bag. Steadi- 

 ness of action can, however, be attained by further adding to 

 the mixture common salt, for which hint I am indebted to 

 Mr. E. Holland. Take, say, 



Chlorate of potash . . 2 Ibs. 

 Manganese oxide .... f Ib. 

 Salt ...... 6 ounces. 



In this mixture the chlorate is to be powdered as well as 

 the other ingredients, and the evolution of gas will answer 

 almost instantly to the lowering of the flame, and thus be 

 under perfect control. This is therefore the mixture I recom- 

 mend : chlorate, 8 parts by weight ; manganese, 2 parts ; salt 

 1^ parts. 



When chlorate is used with manganese alone, it must 

 not be powdered, but used in the rough crystals ; and 1 Ib. of 

 the chlorate the only active ingredient must be allowed 

 for 4 to 4 \ cubic feet of oxygen. Powdered chlorate, without 

 salt, is liable to cause violent rushes. It may also be noted 

 that finely granulated manganese is better than powdered, as 

 not choking the delivery-tube at all ; but it is seldom easy to 

 procure, and the powder need cause no inconvenience. 



Using the necessary caution, it often happens that when 

 the charge is about half exhausted the evolution of gas stops 

 for a while, and even the full flame fails to start it again for 

 some minutes. It will resume, however, and during this 

 second stage there is little need of caution for fear of a rush ; 

 the full heat is generally needed to the end. For its thorough 

 manageability, however, I strongly recommend the salt mix- 

 ture. 



