The British Acetylene Association's regulations relating to carbide of calcium (_vide_ Chap. XIV.) contain a clause to the effect that "carbide which, when properly decomposed, yields acetylene containing from all phosphorus compounds therein more than 0.05 per cent, by volume of phosphoretted hydrogen, may be refused by the buyer." This limit is equivalent to 0.74 milligramme of phosphorus calculated as PH_3 per litre. A latitude of 0.01 per cent, is, however, allowed for the analysis, so that the ultimate limit on which carbide could be rejected is: 0.06 volume per cent. of PH_3, or 0.89 milligramme of phosphorus per litre. The existence in appreciable quantity of combined silicon as a normal impurity in acetylene seems still open to doubt. Calcium carbide frequently contains notable quantities of iron and other silicides; but although these bodies are decomposed by acids, yielding hydrogen silicide, or siliciuretted hydrogen, they are not attacked by plain water. Nevertheless Wolff and Gerard have found hydrogen silicide in crude acetylene, and Lewes looks upon it as a common impurity in small amounts. When it occurs, it is probably derived, as Vigouroux has suggested, from "alloys" of silicon with calcium, magnesium, and aluminium in the carbide. The metallic constituents of these substances would naturally be attacked by water, evolving hydrogen; and the hydrogen, in its nascent state, would probably unite with the liberated silicon to form hydrogen silicide. Many authorities, including Keppeler, have virtually denied that silicon compounds exist in crude acetylene, while the proportion 0.01 per cent. has been given by other writers as the maximum. Caro, however, has stated that the crude gas almost invariably contains silicon, sometimes in very small quantities, but often up to the limit of 0.8 per cent.; the failure of previous investigators to discover it being due to faulty analytical methods. Caro has seen one specimen of (bad) carbide which gave a spontaneously inflammable gas although it contained only traces of phosphine; its inflammability being caused by 2.1 per cent. of hydrogen silicide. Practically speaking, all the foregoing remarks made about phosphine apply equally to hydrogen silicide: it burns to solid silicon oxide (silica) at the burners, is insoluble in water, and is spontaneously inflammable when alone or only slightly diluted, but never occurs in good carbide in sufficient proportion to render the acetylene itself inflammable. According to Caro the silicon may be present both as hydrogen silicide and as silicon "compounds." A high temperature in the generator will favour the production of the latter; an apparatus in which the gas is washed well in lime-water will remove the bulk of the former. Fraenkel has found that magnesium silicide is not decomposed by water or an alkaline solution, but that dilute hydrochloric acid acts upon it and spontaneously inflammable hydrogen silicide results. If it may be assumed