72 LIGHT PRODUCED BY THE AGENCY OF ELECTRICITY. 



considered that oxygen, after having submitted to combustion, 

 having parted with the light with which it was supposed to be 

 combined, was no longer capable of affording light: this is com- 

 pletely contradicted by the following experiment. 



Place in Dr. lire's eudiometer a minute quantity of each of 

 the following gases, having previously submitted them sepa- 

 rately to the electric spark, viz., oxygen, carbonic-oxide, azote, 

 and hydrogen, they will, in their dephlogisticated state (as it is 

 termed), produce light by the agencies of electricity ; proving 

 thereby that it was not previously combined with any of them as 

 a specific body, but that they each contribute to its direct for- 

 mation^ and exhibit the transition of matter to a radiant state. 

 It is not here asserted that the entire of these gases will be at 

 once converted without leaving a residuum of fixed matter, but 

 it is affirmed that the nearer they are proportioned to their 

 relative quantities, the less will be the residual gases or product 

 unconverted. The quantity operated upon being so small, 

 difficulties arise in affording exactitude of proportion, which is 

 also in itself to be yet more perfectly attained by a series of 

 experiments, which require the greatest precision. 



Happily, the durability of fixed matter throughout nature is 

 secured by the difficulty of producing, on an extensive scale, the 

 matter of the three essential rays under the requisite circum- 

 stances of proportion and position, to constitute their immediate 

 conversion to the radiant state. Were it otherwise, it would 

 overturn the beautiful arrangement which exists according to 

 the order of Providence, than which there can be nothing 

 more sublime. 



The conversion to the one state, and the restoration to that 

 of the other, are gradual ; and observe the necessary laws which 

 effect the preservation of a most perfect equilibrium. 



The spontaneous combustion of ammonia in chlorine is 

 mentioned by Dr. Thompson, vol. iii. page 47. He gives the 

 particulars of an experiment in which there appears a defi- 

 ciency unaccounted for, which is cited here as one of the many 



