object of these instruments is to restrict the of gas when superabundant, and it should be noticed that the supply requires to be so before any advantage can result from the use of them. A conical valve, operating by the pressure of the gas in a manner similar to the governor at the gas-works, is in general the acting part of the apparatus.
The impurities which should be removed in the manufacture of coal-gas are sul phureted hydrogen, ammonia, and carbonic acid. The presence of sulphuretcd hydrogen is detected by allowing a stream of the gas to play on a paper wetted with a solu tion of acetate of lead; the test-paper is blackened if the deleterious gas be present. Ammonia is detected by allowing the gas to play on paper stained yellow with turmeric. Ammonia changes the yellow to brown. The presence of carbonic acid can be ascer tained by causing the gas to bubble through lime-water. If carbonic acid be present, it combines with the lime, and the water becomes milky.
The value of gas for lighting depends on its illuminating power, which, again, mainly depends on the proportion of olefiant gas and heavy hydrocarbons contained in the mixture. The specific gravity of the gas would be a complete test of the illumi nating power, were it first ascertained that no were contained in the mixture. The chlorine and bromine tests, which are applied by bringing the gas into contact with either of these, substances in a graduated tube, alSo require that the absence of deleterious gases be ascertained. Chlorine and bromine condense the olefiant gas and heavy hydrocarbons, and the proportion of them present is ascertained by the pro portion of the gas which is condensed.
The most practical mode of determining the illuminating power is by the use of the Bunsen photometer, introduced into this country by Dr. Lyon Mayfair, first adapted by the late Mr. King of Liverpool, and since modified and improved by Dr. Letheby and Mr. F. J. Evans of London. At one end of a straight bar of wood, a gas-burner is mounted; on the other end, a candlestick. These are so placed, that when lighted, there are exactly '100. in. between the centers of the lights. The bar is correctly graduated to show how many times the one light exceeds the cam A circular disk of paper made semi-transparent, excepting a spot in the center, which is left opaque, is placed at right angles across the graduated bar on a stand which slides along the bar.
When the disk is moved into a position where the opaque spot is invisible, the lights are equal—the disappearance of the spot being caused by the light transmitted by the semi transparent part of the disk being equal to that reflected by the opaque part. The figures immediately below the disk indicate the power of the light.
As has been stated, the illuminating power of coal-gas may vary from 12 sperm candles up to nearly 40, though neither extreme,is supplied to consumers. In England, quality varies from 14 to 22 candles; in Scotland, from 22 to 32 candles. The cost of production is affected by causes independent of the illuminating power, and these are so various, that the cost can hardly be the same in any two places. In contrasting the price of gas in - different places, another difficulty arises from the unavoidable varia tion in the quantity accounted for; the loss sustained under the head of condensation, leakage, bad debts, and waste, varies from 10 up to 30 per cent on the whole quantity made; and though, when this loss is excessive, the remedy should, to a certain extent, be in the power of the manufacturer, yet there is a considerable range within which the loss may vary, owing to local and peculiar circumstances which the manufacturer can not control.
The economy of gas for lighting purposes will be apparent when it is considered that 50 ft. of gas consumed in a burner at 5 ft. per hour, will last 10 hours; while a sperm candle of six to the pound, and burning 120 grains per hour, will only last 9.722 hours. Assuming, however, that both will last 10 hours—a view which is in favor of the candle —1000 cubic ft. will last as long as 20 candles; therefore, with an illuminatingpower of 15 candles, it will give an amount of light equal to 800 candles, or 50 lbs., at 2s., per lb. would cost .-e5; at 20 candles it would equal 400, or 661 lbs., costing a 13s. 4d.; at 25 candles it would equal 500. or 83i lbs., costing £8 Gs. 8d.; at 30 candles, it would equal 600, or 100 lbs., costing £10.