OF THE MANGFACTUKE OF GAS. 



91 



gas operated upon, since a large holder, even when depressed 

 to the greatest extent, must always contain a considerable 

 quantity of gas from the previous operation, and thus the 

 experiment is vitiated ; whilst with a smaller vessel, this 

 residue can always be got rid of by allowing a few cubic feet 

 of the gas at the time generating, to blow through it simul- 

 taneously with the collection of the first portion of the 

 sample. 



A rather low heat was employed in all the experiments, as 

 it was found to be the best as well for the coals alone as with 

 water gas ; and the results obtained in the trials of the coals 

 without water gas, will rarely be found below those of other 

 experimenters. 



The temperature of the gases on reaching the meter was 

 found to be no higher than that of the external atmosphere. 

 The greatest care was taken to secure accuracy in the results 

 and perfect fairness in the comparison between the coals 

 distilled alone and with water gas. All the weighings were 

 made before me, and every experiment from beginning to 

 end was made under my own personal inspection. 



The illuminating power was tested by Bunsen's Photo- 

 meter, — a large number of the experiments being made with 

 an improved form of the instrument, invented by Messrs. 

 Church and Mann, of the City Gas Works, London. In 

 some instances, the shadow test was also tried. The size of 

 burner and pressure of gas were in most cases noted, and in 

 every instance the determination of the illuminating power 

 was made when the gas was burning to the greatest advantage, 

 that is, without a flickering flame or a tendency to smoke. 

 These experiments are, however, even with the greatest 

 care, subject to certain errors, caused principally by the irre- 

 gular burning of the spermaceti candle, rendering them only 

 approximative. The liability to these errors has, it is true, 

 been much reduced by the ingenious plan of substituting a 



