90 Dr. Thomson on the Composition of [Aug. 



tube by means of a little amianthus with which I stuffed its 

 extremity. 



Such is the last state to which I have reduced my apparatus. 

 But I find myself still unable to prevent considerable differences 

 from taking place, both in the quantity of gas evolved, and of 

 moisture absorbed by the muriate of lime, in different trials on 

 the same coal. To what these differences are to be ascribed, it 

 is not easy to say. I have been suspecting that the column of 

 muriate of lime through which the gas passes (about 12 inches) 

 is not, perhaps, long enough to deprive it effectually of all its 

 moisture. I intend, therefore, to make trial of a column twice 

 that length to see whether it will yield more steadv results ; but 

 I suspect that the chief, if not the only cause of these differences 

 is the difficulty of reducing the coai to a powder sufficiently 

 fine, and of mixing it with perfect regularity through so great a 

 proportion of peroxide of copper ; for unless every particle of 

 coal be very smaU, and unless it be in contact with a sufficient 

 number of particles of peroxide of copper, it is very possible that 

 it may escape complete combustion. 



In such experiments, therefore, it would be extremely hazard- 

 ous to trust to a single trial. I generally made six successive 

 experiments upon each kind of coal, and sometimes even more. 

 The conclusions were drawn from a mean of all the experiments, 

 except when some obvious reason presented itself for excluding 

 any individual experiment. 



1. Caking Coal. 



From one grain of caking coal, treated in the way above 

 described, there were evolved 5D17 cubic inches of gas, suppos- 

 ing the barometer to stand at3(J inches, and the thermometer to 

 stand at 60°. Of this gas, caustic potash absorbed 5-3167 

 cubic inches, which were considered to be carbonic acid gas. 

 The residual gas (measuring 1T2 cubic inch) was left for 24 

 hours in contact with a stick of phosphorus, at a temperature 

 which varied from 60° to 80° according to the time of the day. 

 The 1 12 volumes of gas by this treatment were reduced to 98-56' 

 volumes. The 13-44 volumes which disappeared, I considered 

 as oxvgen gas. I suppose this oxygen gas to be derived from 

 th'J'common air which occupies the interstices of the peroxide of 

 copper and muriate of lime in the tubes. The amount of this 

 common air is easily estimated when we know r the volume of its 

 oxvgen ; for if we multiply the volume of oxygen by five, we 

 obtain the volume of common air. From this it is obvious that 

 the portion of the residual gas, which was common air, amounted 

 to 0*672 cubic inch. The remainder, amounting to U-448 cubic 

 inch, must have been azotic gas obviously derived from the 

 decomposition of the grain of caking coal on which the experi- 

 ment was made. 



It is obvious that the place of 0-672 cubic inch of common 



