The retorts best calculated for large gas-light works are seven or eight foot long (without the mouth-piece) and twelve inches in diameter, tapering down to ten inches--if they are larger the coal which they contain cannot be heated properly. The advantages that may be derived from the circumstances before stated are of greater value in the gas-light manufacture than is often imagined, and the quantity as well as the quality of the gas is very much influenced by such circumstances. If coal be distilled with a very low red heat scarcely observable by daylight, the gas produced gives a feeble light--if the temperature be increased so that the distillatory vessel is of a dull redness, the light is more brilliant and of a better colour--if a bright or cherry-red heat be employed the gas produced, burns with a brilliant white flame, and if the heat be increased so far that the retort is almost white hot, and consequently in danger of melting, the gas given out, has little illuminating power, and burns with a clear blueish flame;[19] or if the coal abounds in pyrites or sulphuret of iron, as is sometimes the case with Newcastle coal, a large quantity of sulphuretted hidrogen is likewise evolved, which although it increases the illuminating power of the coal-gas, has the capital disadvantage, of producing an intolerable suffocating odour, when the gas is burnt which is particularly perceptible in low rooms illuminated with such gas. [19] It is chiefly a mixture of carbonic oxid, and hydrogen gas. These observations also apply to the distillation of tar, which when distilled either in a vaporous or nascent state, during its first production from coal in the ordinary process, or if it be submitted to a second distillation, mingled with a fresh portion of pit-coal, a practice usually had recourse to when this product cannot be disposed of more advantageously. The best depth of coal in the retort for procuring excellent gas, and at the same time for yielding the greatest quantity from the same weight in the shortest possible time, is about six inches. The brightness of the coal-gas flame is rather diminished when the gas has been long kept over water, and hence for illumination it should be used as soon as prepared, but of course properly purified. The quantity of gas taken up by water is affected by temperature, because the temperature increases its elasticity; the quantity of gas absorbed, diminishes as the temperature increases, and increases as the temperature diminishes. ½7 part of its own bulk of pure coal-gas is absorbed by the water over which it is confined in the gazometer.