upon the Composition of Coal. 125 



this might be the cause of the error. Bichromate of potash 

 and dichromate of lead were then successively employed in- 

 stead of oxide of copper, but it was found impossible to avoid 

 the formation of deutoxide of azote with these substances, in 

 such quantities as to be reduced by the copper. Oxide of 

 lead was also used, but the combustion was quite imperfect, 

 and the quantity of ammonia formed, very great. Oxide of 

 copper strongly heated, was at last employed, but the error 

 remained the same. 



The method remains thus at the present time, but I hope, 

 by further investigation, to discover the cause of the error*. 

 The error being thus confined to 1 per cent., two analyses of 

 coal were made, in order to obtain some idea respecting the 

 amount of the azote. These coals will afterwards be more 

 minutely described. 



I. '283 grm. Coal from the neighbourhood of Edinburgh. 



27°7 Barometer. 



is"* Thermometer. 



11'8 cc ... Air in the receiver. 



12'7 cc ... Gas and air after the experiment. 



•9 cc ... Azote. 

 •9 cc Azote reduced to 0° thermometer and 28° barometer give 

 •84? cc which equals '38 per cent. 



II. '300 grm. Coal from Garesfield near Newcastle-on- 

 Tyne, produced no gas, but there appeared to be a diminu- 

 tion of '30 cc. 



From the above experiments it clearly appears, that the 

 coal cannot contain more than 2 per cent, of azote, but with 

 the present means of analysis at our disposal it is impossible 

 to determine the true amount. 



II. ANALYSES OF THE COALS, &C. 



The arrangement of the various coals, proposed by Dr. 

 Thomson, has been for the present adopted in the following 

 account. Two specimens of each of these varieties from dif- 

 ferent localities have been analysed. There are four varieties, 

 viz.. Splint, Cannel, Cherry, and Caking. 



I. VARIETY — SPLINT COAL. 



1. Specimen from Wylam Banks. 

 This coal is not at present worked ; it is a thin bed very 

 low down in the Newcastle coal series, and appears in this spot 



* Since the above was first printed, it has occurred to me that the error 

 arises from the absorption of the oxygen gas of the air in the apparatus, 

 by the reduced oxide of copper, after the analysis is finished and during the 

 cooling of the tube of combustion. 



