having been scraped off, the plate is removed from the table and placed for several days in a more inclined position, to complete the drainage of the mercury.

Plate-glass is not always fashioned, however, by casting. The process of blowing a plate of this kind of glass is very similar to the mode of blowing spread or broad glass, described at page 55; the minute differences observable in the manipulations in these processes being chiefly occasioned by the weight and bulk of the mass of glass operated on being considerably greater in the blowing of plate-glass than in the blowing of spread-glass. It is very difficult to blow a plate of glass of sufficient thickness for grinding to a level surface and polishing, of a size larger than fifty inches in length by forty inches in breadth. By the process of casting, plates of glass are produced of one hundred and sixty inches in length by eighty inches in breadth.

The flutings and triangular indentations which are sometimes seen on one side of plate-glass are produced by corresponding but reverse flutings and indentations on the surface of the heavy metallic roller with which the liquid glass is spread out upon the casting-table.

V. DEFECTS IN GLASS.

In

The attainment of a process for preparing perfect glass has for a long time, and with only doubtful success, engaged the attention of the experimental philosopher, as well as that of the glass manufacturer. The most important defects likely to exist in glass are striæ or wreath, knots, threads, and tears. Striæ are undulating appearances which result from a want of uniformity in the density of the mass. consequence of this inequality, the rays of light passing through the glass become differently refracted and dispersed, thus producing a wavy appearance like what is perceived in looking through a liquid composed of two imperfectly mixed solutions of dif ferent strengths. This defect may exist to a considerable extent without being discernible by the naked eye, and consequently without being detrimental to the quality of the glass when applied to all ordinary purposes; but the smallest incongruity in the mass would render the glass quite inapplicable to the construction of optical instruments, where every distortion of the rays of light in their passage through the glass would be considerably magnified. Flint-glass is much more subject to this defect than any other kind of glass. The opaque knots which are sometimes perceived in glass may be either particles of earthy matter fallen from the furnace or abraded from the crucible, a portion of the glassgall, or else imperfectly vitrified grains of sand. The appearances called threads and tears result from pieces of partially vitrified clay, which fell into the crucible

from the roof of the furnace, but did not mix uniformly with the glass. These semi-vitreous lumps or threads differ from the proper glass not only in colour, but in their degree of dilatation and contraction through change of temperature, which gives the glass a disposition to fly to pieces without any apparent cause. Another defect sometimes perceived in the less fusible varieties of glass is a want of clearness, owing to the presence of minute bubbles diffused throughout the substance of the glass. The presence of these bubbles indicates that the glass has not been maintained in a perfectly fluid state for a sufficient length of time to allow of their union and dispersion.

With the exception of the striæ, these defects may be avoided without any extraordinary care and attention; and in the case of crown-glass, where the specific gravities of the materials do not greatly vary, and where the temperature at which the materials are vitrified is very intense, defects arising from the existence of striæ are not of great importance. But in the case of flint-glass, the materials for which have a very unequal density, it is extremely difficult to obtain, by the usual process, a perfectly uniform piece large enough to form a disc of only a few inches in diameter; and the fitness of glass for being constructed into lenses depends much more on its uniformity in density or refractive power, than on its whiteness or brilliancy. If the raw materials have not been well incorporated, the pot of glass is very apt to contain an excess of oxide of lead in some parts and a deficiency in others, which pro

duces a corresponding inequality in the density of the mass; and even in the slow refrigeration of a pot of pretty uniform glass, the heavier silicate of lead always subsides, and leaves the lighter alkaline silicate in excess at the upper parts; but if the mass is quickly cooled, it would be extremely brittle, and apt to fly to pieces on attempting to cut it.

The want of uniformity in a pot of flint-glass seems to proceed in a great measure from the unequal heating of the pot, the construction of the furnace being generally such as to allow the sides to be heated considerably higher than the bottom. When this is the case, the denser portions of glass, which subside to the bottom, remain there; but if the temperature of the inferior part of the pot is made considerably higher than that of the upper part, the denser glass would become expanded by the increased heat, and probably be diffused throughout the mass. The figure of a furnace said to be adapted to this object may be found in Dr. Ure's Dictionary of Arts, &c. art. Glass-making.

Another method of obtaining a pretty uniform pot of glass is by agitation, the glass being either poured by ladles from one crucible into another in the furnace, or else stirred by a rod of iron encased in stoneware, as the naked iron itself would colour the glass. After being agitated, the glass must be allowed to stand quiet for a short time for the airbubbles to be dissipated; and when perceived by the proof to be perfectly "fine" or free from bubbles, it should either be quickly cooled while in the cru

* This mode of stirring was adopted by the late M. Guinand.