PREPARATION OF GASES 75 



the latter, which only acts mechanically, some use fine 

 washed sand, or even red oxide of iron, to avoid a possible 

 danger lurking in the black colour of the manganese, which 

 has been known to be adulterated with powdered charcoal. 

 Should this be the case, the carbon would probably ignite 

 with the heat, and, burning in the evolving oxygen with 

 explosive violence, produce what may be called a carbon ex- 

 plosion in the retort. A similar explosion once occurred when 

 the manganese had become, not adulterated, but accidentally 

 mixed, with antimony sulphide. From a known source, no 

 dread need be felt ; but in case of doubt a drachm of the 

 mixture should be placed in a dry test-tube and heated over 

 a Bunsen burner. It will always crackle, and a few tiny sparks 

 may be seen ; but if large, and bright sparks appear, or anything 

 like a small explosion be heard, it should be rejected. 



Any sand or iron oxide, if these are substituted, should 

 be known to be perfectly cleaned. Of the chlorate itself, the 

 common commercial quality is quite pure enough ; but for 

 similar reasons, every parcel purchased should be spread out 

 upon a large sheet of the Times, and any bits of wood or 

 straw carefully picked out, should any such be visible. 



40. The Oxygen Retort. The mixture of which imme- 

 diately must be placed in a retort, of which three kinds are 

 sold. First, a copper one, the dearest to buy and the worst 

 to wear of any. Secondly, a conical flat-bottomed one of 

 sheet iron (fig. 46), which is perhaps the most usual, and fairly 

 durable. It should have a safety-tube at the top, in which a 

 cork can be placed, as well as the deli very- tube. This latter 

 should be of a good diameter, as a certain quantity of the 

 powdered manganese is carried up into it by the rapidly evolv- 

 ing gas, and a small tube might choke, when too great a rush 

 of gas might produce a sort of mild confined pressure explosion. 

 The cork in the safety-tube will sufficiently guard against this, 

 blowing out if necessary and letting the gas escape ; and there 

 is no fear if the delivery-tube is examined after every operation, 



