Devitrification of Glass. Reaumur's Porcelain.—If glass, surrounded by sand or gypsum, be heated for a considerable time, but not sufficiently to cause fusion, it will be converted into a porcelain-like mass. This result may also be obtained, but with less regularity, by melting an impure glass, such as ordinary bottle-glass, in a crucible in a furnace, and allowing the furnace to die out. This change of condition is supposed to be caused by the partial separation of certain silicates, especially those of calcium and aluminium, and their assumption of a more or less crystalline form. A mass of these crystals, analyzed by Dumas, gave the formula ; whilst the transparent glass from whioh they separated contained 3.5 per cent, less silica, less alumina, and a propor tionately larger quantity of soda. It has been observed that devitrified glass may be vitrified and again devitrified ; that from devitrified glass, placed in a moist atmosphere, there exudes a soluble salt ; and that apparently the melting-point of devitrified glass is higher than that of the same glass when vitrified. Glass of any kind may be devitrified ; but the finer kinds of potash glass only with difficulty. The soluble glass of Fuch is especially liable to devitrifaction by crystallization. Attempts have been made to utilize devitrified glass, but without any marked success.
Badly prepared window-glass becomes opaque on exposure to air. This is due to a devitrifica tion of the surface, brought about by the action of water, carbonic acid, and ammonia, and depends on the fact that the alkalies are separated, and washed away by water, whilst an iridescent coating, consisting of a thin film of caleic silicate, remains on the surface.
RAW MATERIALEI.—Sand.—Freedom from colour, and transparency, in glass, depend principally upon the quality of the sand used in its composition. The impurities generally present in sand are iron, lime, alumina, and magnesia. Iron is the most detrimental, and the quality of a sand is mainly determined by the quantity of iron contained in it.
Sand of extreme purity has been shipped from America, but that generally used in England is obtained either from Alum Bay, in the Isle of Wight, or from the Forest of Fontainebleau, in,' France. Fontainebleau sand is obtained in blocks from the quarries in the same manner as ordinary sandstone. The following are analyses of samples of sand from Alum Bay and Fontaine bleau, and may be taken as typical of sands suited to the production of the best forms of glass :— Alum Bay : silica, 97 per cent.; alumina, magnesia, and oxide of iron, 2 ; moisture, 1. Fontaine bleau: silica, 98.8; magnesia and oxide of iron, 0.7 ; moisture, 0.5.
Before admixture with the other ingredients, the sand is washed and burned. The process of Washing may remove a certain quantity of extraneous dirt and chalk, but it is doubtful whether it rewards the labour expended. Burning tends to disintegrate the larger lumps, and removes all
moisture and organic matter. The removal of organic matter is important, as its presence would tend to reduce a portion of red-lead to the metallic state, or to convert ferric into ferrous oxide. Burning is carried on in a special oven, in which the sand rests upon a fire-clay bed, shelving towards a central trap-door, communicating with a vault beneath. The flame playa over the sand from a fire-place on one side, towards the flue, which is placed on the opposite side to the fire-place, and in such a position that all the sand shall be subjected to the heat. When the sand has become incandescent, the trap-door is drawn, and the sand falls into the vault below. The sand is sifted, both before washing and after burning, through:very fine copperrgauze sieves; it should be kept from contact with iron, and be moved only with wooden shovels.
Tests.—The best test for sand is microscopic examination. Pure sand should be perfectly white, and should not effervesce, nor lose colour, when heated with an acid ; effervescence shows the presence of chalk; oxide of iron would be dissolved by boiling *itb hydrochloric acid, and be discovered by the ferrocyanide test. Pure sand is insoluble in all acids except hydrofluoric.
In the manufacture of common wine and beer bottles, the effect of iron is clearly seen ; their dark-green tint is due to ferrous oxide, present as an impurity in the sand and some of the other materials. Even iu the best samples of sand, there is nearly always a trace of iron. The in jurious colouring effect of small quantities can be neutralized by the use of dioxide of manganese, or of trioxide of arsenic. A green colour is due to ferrous oxide ; if this be converted into ferric oxide, the green becomes a pale-lemon tint, invisible when the quantity of iron is small. The oxidation of ferrous oxide is accomplished through the decomposition by heat of manganese dioxide. If the latter be added in excess, or if the heat be not continued sufficiently long for its decom position, the glass acquires a pink tint, and is said to be high-coloured. From fear of producing high-coloured glass, through using excess of manganese dioxide, and from the fact that glass containing it becomes high-coloured by the continued action of sunlight, manufacturers of window glass prefer arsenic trioxide. If this be heated with ferrous oxide, metallic arsenic volatilizes, leaving its oxygen to convert the ferrous into ferric oxide. When manganese dioxide is used, it should be specially prepared by precipitation, as the natural oxides always contain iron.