104 Dr. Thomson on [August, 



filter. The liquid that passes through has a yellow colour, and 

 a slightly acidulous taste. It contains no traces of prussic acid, 

 nor of ammonia; but is merely a solution of bisulphate of potash 

 'and persulphate of iron. 



The gas thus obtained has a peculiar smell, not disagreeable, 

 nor strong ; but which I can compare to nothing else. Its taste 

 is rather aromatic, and it leaves a hot impression in the mouth, 

 which continues for some time. It is not altered by standing 

 over water exposed to the light. Its specific gravity is Q'993, 

 that of common air being reckoned TOuO. It burns readily when 

 brought in contact with the flame of a candle, and the flame 

 has a deep blue colour. It detonates readily by the electric 

 spark when mixed with oxygen gas. I made a good many trials 

 this way in order to determine the proportion of oxygen neces- 

 sary to consume it completely, the change of bulk that takes 

 place after the explosion, and the new products formed. These 

 experiments were made partly over water and partly over mer- 

 cury. The results were so simple that they will be readily 

 understood without transcribing a tabular view of the trials, 

 which were 12 in number. 



Three volumes of the gas require for complete combustion two 

 volumes of oxygen gas. The residue after the detonation, is 

 three volumes, which consists entirely of carbonic acid gas. 

 Thus on complete combustion the whole oxygen disappears, and 

 the bulk of the inflammable gas is not altered ; but it is totally 

 converted into carbonic acid gas. 



From these facts, it is evident that the inflammable gas is a 

 compound of 



3 volumes carbonic oxide ? , , • , ,, , 



t , i > condensed into three volumes. 



1 volume hydrogen gas 3 



One and a half volume of the oxygen is employed in converting 

 the carbonic oxide into carbonic acid. The remaining half 

 volume of oxygen combines with the volume of hydrogen, and 

 is converted into water. The specific gravity of such a com- 

 pound is by calculation 0-995. Now this almost agrees with 

 that of our gas, which was found experimentally to be 0*993. 



This gas possessing peculiar properties must be considered as 

 a new species. From its composition it may be denominated 

 hydrogureted carbonic oxide. The result of the analysis of this 

 gas occasioned some surprize at first ; for I was not prepared to 

 expect a gaseous compound of hydrogen and carbonic oxide. 

 But when we consider the vast number of vegetable bodies 

 composed of oxygen, carbon, and hydrogen, we cannot hesitate 

 to admit that these substances are capable of uniting with each 

 other in a great many proportions. The great variety of inflam- 

 mable gases obtained during the analysis of vegetable bodies and 

 coal, is but imperfectly explained by supposing it composed of 

 mixtures of the three gases already known. This is probably 



