99 



one anollier ? This would, at least, account for the fact, that gases 

 can scarcely become luminous or lose heat by radiation. 



But, to return from such uncertain ground, the purity of the 

 vgases must be attended to. I procure my hydrogen from frag- 

 ments of zinc ; the oxygen from peroxide of manganese contains 

 nitrogen, so nearly | of its volume whenever I examined it, that I 

 am inclined to suspect it is chemically combined, which is also pro- 

 bable from the circumstance that the water whicli is contained 

 in the native oxide, and is disengaged during its decomposition, 

 is acidulous with nitric acid. If it be used, it should be care- 

 fully dried, as, if it has been exposed to a damp atmosphere, it 

 attracts enough of moisture to produce explosions in the process. 

 I have seen such an accident. Chlorate of potash affords oxy- 

 gen nearly pure, but its high price precludes its use; red pre- 

 cipitate will be found most convenient and economical ; it is ea- 

 sily decomposed in a glazed earthen retort, which will serve 

 repeatedly, and the value of the revived mercury is more than 

 half the price of the oxide. The gas is most impiu'e in the 

 beginning of the distillation ; its impurity has been with me from 

 -J to J5-, rejecting the first bubbles. This detail is minute, but 

 it will, I trust, be /nuiid useful, as the heat depends materially 

 on the quality of the gas ; in fact, in a series of experiments the 

 purity of it should be always stated, and the process of ana- 

 lysis by Volta's Eudiometer is so easy that it may be performed in 

 a few minutes. 



I have tried many experiments with this apparatus, and can 

 say that it is a valuable addition to the laboratory ; to describe 

 them at length would be useless, as most of them were but 

 repetitions of what others had done. That the earths are re- 



p2 



