for colour-matching, in which the latter is known almost to equal full daylight, and to excel every form of artificial light except that of the electric arc specially corrected by means of glass tinted with copper salts. CONDITIONS FOR INCANDESCENT ACETYLENE LIGHTING.--For success in the combustion of acetylene on the incandescent system, however, several points have to be observed. First, the gas must be delivered at a strictly constant pressure to the burner, and at one which exceeds a certain limit, ranging with different types and different sizes of burner from 2 to 4 or 5 inches of water. (The authors examined, as long ago as 1903, an incandescent burner of German construction claimed to work at a pressure of 1.5 inches, which it was almost impossible to induce to fire back to the jets however slowly the cock was manipulated, provided the pressure of the gas was maintained well above the point specified. But ordinarily a pressure of about 4 inches is used with incandescent acetylene burners.) Secondly, it is necessary that the acetylene shall at all times be free from appreciable admixture with air, even 0.5 per cent, being highly objectionable according to Caro; so that generators introducing any noteworthy amount of air into the holder each time their decomposing chambers are opened for recharging are not suitable for employment when incandescent burners are contemplated. The reason for this will be more apparent later on, but it depends on the obvious fact that if the acetylene already contains an appreciable proportion of air, when a further quantity is admitted at the burner inlets, the gaseous mixture contains a higher percentage of oxygen than is suited to the size and design of the burner, so that flashing back to the injector jets is imminent at any moment, and may be determined by the slightest fluctuation in pressure--if, indeed, the flame will remain at the proper spot for combustion at all. Thirdly, the fact that the acetylene which is to be consumed under the mantle must be most rigorously purified from phosphorus compounds has been mentioned in Chapter V. Impure acetylene will often destroy a mantle in two or three hours; but with highly purified gas the average life of a mantle may be taken, according to Giro, at 500 or 600 hours. It is safer, however, to assume a rather shorter average life, say 300 to 400 burning hours. Fourthly, owing to the higher pressure at which acetylene must be delivered to an incandescent burner and to the higher temperature of the acetylene flame in comparison with coal-gas, a mantle good enough to give satisfactory results with the latter does not of necessity answer with acetylene; in fact, the authors have found that English Welsbach coal-gas mantles of the small sizes required by incandescent acetylene burners are not competent to last for more than a very few hours, although, in identical conditions, mantles prepared specially for use with acetylene have proved durable.