.s7vY \VIU.1AM SIEMENS, F.R.S. 339 



The usr of \\ivter-gaa, resulting from the decomposition, of steam 

 in passing through a hot chamber filled with coke, has been 

 suggested, but this gas also is objectionable, because it contains, 

 1 -sides hydrogen, the poisonous and inodorous gas carbonic oxide, 

 tin- introduction of which into dwelling-houses could not be 

 effected without considerable danger. A more satisfactory mode 

 of supplying heating separately from illuminating gas would 

 consist in connecting the retort at different periods of the 

 distillation with two separate systems of mains for the delivery 

 of the respective gases, as has been proposed by me elsewhere. 

 Experiments made some years ago by Mr. Ellisen of the Paris 

 gasworks have shown that the gases rich in carbon, such as 

 olefiant and acetylene, are developed chiefly during an interval 

 of time, beginning half an hour after the commencement and 

 terminating at half the whole period of distillation, whilst, during 

 the remainder of the time, marsh gas and hydrogen are chiefly 

 developed, which, while possessing little illuminating power, are 

 most advantageous for heating purposes. By resorting to improved 

 means of heating the retorts with gaseous fuel, such as have been 

 in use at the Paris gasworks for a considerable number of years, 

 the length of time for effecting each distillation may be shortened 

 from six hours, the usual period in former years, to four, or even 

 three hours, as now practised at Glasgow and elsewhere. By this 

 means a given number of retorts can be made to produce, in 

 addition to the former quantity of illuminating gas of superior 

 quality, a similar quantity of heating gas, resulting in a diminished 

 cost of production and an increased supply of the valuable by- 

 products previously referred to. The quantity of both ammonia 

 and heating gas may be further increased by the simple expedient 

 of passing a streamlet of steam through the heated retorts towards 

 the end of each operation, whereby the ammonia and hydrocarbons 

 still occluded in the heated coke will be evolved, and the volume 

 of heating gas produced be augmented by the products of decom- 

 position of the steam itself. It has been shown that gas may be 

 used advantageously for domestic purposes with judicious manage- 

 ment even under present conditions, and it is easy to conceive 

 that its consumption for heating would soon increase, perhaps 

 tenfold, if supplied separately at, say, Is. a thousand cubic feet. 

 At this price gas would be not only the cleanest and most 



z 2 



