142 Analj/ses of Books, [Aug. 



following table of the comparative expence of lighting with these 

 two gases, and with oil and tallow. 



s. d. 

 Valuing the quantity of light which 1 lb. of tallow gives 



in candles at 1 



An equal quantity of light from sperm, oil consumed in 



an Argand*8 lamp, will cost 6-?- 



Ditto from whale oil gas . , 4^ 



Ditto from coal gas • 2|- 



"Twenty cubic feet of coal gas, or ten of oil gas, he considers 

 as equivalent to a pound of tallow, and 5000 grains of good 

 sperm, oil to 7000 of tallow, or 1 lb. avoirdupois. 



" The advantages of oil gas over gas from coal are, that smaller 

 distilling vessels are required ; that gasometers and conduit pipes 

 of half the capacity are sufficient ; that no washing apparatus is 

 necessary ; that the trouble and expence of removing waste 

 materials is avoided ; and that the gas affords a much brighter 

 liffht, and with a smaller production of heat, and also of water. 

 When only a moderate quantity of light is required ; when it is 

 an object to save room or labour; and in countries where coal is 

 dear, oil gas is entitled to a decided preference ; but it cannot 

 be brought into competition with coal gas, where coal is cheap, 

 or where the establishments to be lighted are of very considerable 

 magnitude, and of such a nature as to allow of their being freely 

 ventilated. 



** Of the comparative value of different compounds of hydro- 

 gen and charcoal for the purpose of illumination, it still appears 

 to me that the only accurate test is the one which I proposed in 

 Nicholson^s Journal for 1805, viz. the quantities of oxygen gas 

 required to saturate equal volumes. If 100 measures, for 

 instance, of one gas, require for perfect combustion 100 measures 

 of oxygen, and 100 measures of another gas take 200 of oxygen, 

 the value of the second will be double that of the first. Specific 

 gravity, though a guide to a certain extent, is not a sufficient 

 one, for the weight of a gas may be owing to a large proportion 

 of carbonic oxide, which is capable of giving out only a very 

 small quantity of light. Photometrical experiments also appear 

 to me to require greater perfection in the instruments that have 

 been invented for that purpose, before we can implicitly trust to 

 results obtained by their means ; but there can be no fallacy in 

 the combustion of these gases by oxygen, if conducted with ordi- 

 nary care, and especially if, in each instance, an average be 

 taken of two or three trials, which need not occupy more than a 

 few minutes. Nor can it admit of a doubt that, other circum- 

 stances being equal, the brilliancy of light evolved by the com- 

 bustion of gases which are constituted of purely inflammable 

 matter, will bear a proportion to their densities, perhaps even a 



