truth, and thus, as already said, to avoid the necessity of conducting long and uncertain preliminary experiments.

The specific gravity of glass is influenced by the degree of heat to which it has been exposed during its vitrification; being always least when the temperature has been greatest. The cause of this variation is to be found in the different quantities of alkali that have been dissipated, the silica appearing to depart from the completeness of its aggregation, in proportion as it is deprived of alkali.*

Flint glass is not entitled to any drawback upon its exportation, unless its specific gravity be at least three times that of water. The duty drawn back on the shipment of flint glass is considerably greater than the rate originally paid on other descriptions; and but for the resort which is had to its gravity as a test, those other descriptions might in many cases be substituted, and a considerable profit be thereby fraudulently obtained by the exporter at the expense of the revenue.

* The rule for ascertaining the mean specific gravity of different bodies should never be relied on until verified by experiment The condensation of volume which some substances undergo when brought into combination with others, is such as to render all calculations concerning them, under such circumstances, vague and erroneous. From the experiments of M. Loysel it would appear, however, that except in the case which he has noticed, that of employing different degrees of heat, this condensation does not occur with the various compositions of which glass is formed, and recourse may therefore be had to the formula usually employed for ascertaining mean specific gravities, when we desire to determine those of dif ferent vitreous combinations. The rule is as follows:

The specific gravity of one body is to that of another, as the weight of the first divided by its volume is to the weight of the second divided by its volume; and the mean specific gravity of the two is found by dividing the sum of the weights by the sum of the volumes.

Let W, w, be the two weights; V, v, the two volumes; P, p, the two specific gravities; and M, the calculated mean specific gravity.

CHAP. XIII.

ON THE ART OF COLOURING GLASS.

ANTIQUITY OF THIS ART. SPECIMENS OF ROMAN MOSAIC.ANALYSIS OF THESE BY KLAPROTH.-METALLIC OXIDES.

GOLD-PURPLE.-ITS GREAT COLOURING POWER. KUNCKEL.HIS PROFICIENCY IN COLOURING GLASS. - YELLOW COLOUR FROM SILVER FROM LEAD - FROM TARTAR FROM BEECHWOOD CHARCOAL - FROM OXIDE OF IRON. GREEN. BLACK GLASS. BLUE. DIRECTIONS FOUND IN OLD AUTHORS. -IMITATION OF THE GARNET OF THE AMETHYST - OF THE EMERALD OF SAPPHIRES. OPAQUE GLASSES - BLACK OPAL. ANCIENT PICTURES FORMED OF COLOURED GLASS HOW EXECUTED. DESCRIPTION OF ANCIENT MOSAICS. MORE RECENT PROSECUTION OF THIS ART. ACCIDENTAL COLOURING OF PLATE GLASS AT ST. GOBAIN. INEFFECTUAL ATTEMPTS TO REPRODUCE THIS EFFECT.

WHITE.

Ir appears probable that the art of colouring glass was discovered and prosecuted at a period very little subsequent to that of the manufacture of the article itself. The most ancient authors who have mentioned the existence of the material, have also recorded the fact of its being tinged with various colours, in imitation of gems. Strabo, Seneca, and Pliny, all make mention of this use, as being one to which glass was applied by artists in very early times.

The fact has already been mentioned of coloured figures having been found with Egyptian mummies, and which are, therefore, known to have been in existence for upwards of 3000 years. These curious relics of ancient times have also been discovered decorated with coloured glass beads; and a mummy thus ornamented is to be seen in the British Museum.

In the reign of Augustus, the Romans began the use of coloured glass in the composition of mosaic decorations. Several specimens of this kind have been found at a late

period, among the ruins of a villa built by Tiberius in the island of Capri; and some of these specimens having been subjected to analysis by the accurate and ingenious Klaproth, it is known that in that early time recourse was had to the same class of colouring ingredients as is employed by the moderns. Some difference must, indeed, have been observed in their processes, as the ancients were unacquainted with the use of the mineral acids, which are now found to be so convenient in the preparation of metallic oxides.

Klaproth has given the following as the result of his examination of some of the Roman specimens above mentioned:

One which was a lively copper red, opaque, and very bright where recently fractured, contained, in 200

grains, Silex

Oxide of lead

copper

iron

Alumine

Lime

142

28

15

2

5

3

195.

Another, a light verdigris green, also opaque, with a splendent fracture and scoriaceous, contained in a similar quantity,

Silex

Oxide of copper

lead

iron

Lime
Alumine

130

20

15

7

13

11

196.

It is remarkable that the constituent ingredients of both these specimens should prove to be the same.

The difference between then exists only in their relative proportions; and the colours depend upon the different degrees of oxidation of the copper. Sub-oxide of copper,that is, copper which has combined with itself only half the quantity of oxygen required for the production of the perfect oxide — produces a red enamel; while that which has received its full proportion of oxygen yields a green enamel colour.

The specimen of ancient blue glass which was analysed by Klaproth contained,

It appears, therefore, to have been indebted to the oxide of iron for its blue colour, as no trace was detected of any other ingredient to which this could be referred. Since the discovery of the certain and commodious method of producing blue enamel by means of cobalt, the art of obtaining this colour from iron has been lost.

The causes which influence the employment of metallic oxides for the embellishment of porcelain have been sufficiently detailed in the preceding treatise. The same reasons also oblige the artist to have recourse to the same class of substances for imparting colours to glass. The mode of application of colouring materials to these two branches of manufacture differs, however, in this, — that while, in ornamenting porcelain, they are applied superficially, in the manner of pigments, they enter more intimately into the composition of glass, being transfused through the whole mass, and equally incorporated with its entire substance.

The preparation of metallic oxides as colouring materials is nearly similar in all cases; it will not, therefore,

be necessary here to repeat directions, or to give many explanations upon that head.

Gold, in a state of great division and oxidated, has long been celebrated as a means for imparting to glass a most exquisite purple-red colour resembling the ruby, and nearly equalling that gem in the richness of its hue. It is not by any means easy to prepare glass of this colour with any certainty of a successful result. The great tendency which is shown by gold to assume the reguline state, when exposed to excessive heat, to carbonaceous vapours, or to the action of hydrogen, renders necessary a great degree of careful management in the various processes.

The manner has already been given of preparing the purple precipitate of Cassius; the form wherein gold has been used with so much celebrity in imparting red and purple colours. It has been very generally imagined, that the tin used in the preparation of this precipitate is essential to the production of the requisite colour; an opinion which has been shown to be void of foundation, as preparations of gold have been made without the agency of tin, and which have equally possessed the power of imparting the finest purple colour to gold. The colouring property of any of the simple oxides of gold is found to be materially greater than that of Cassius's precipitate; which circumstance has been brought forward as another argument to prove that the presence of tin is far from adding any thing to the body of the colour. It is probable, however, that, although not indispensable, tin is yet useful, as enabling the gold to bear without reduction a higher degree and a longer continuance of heat. With this same object, it has been recommended to add to the precipitate, before using it, a small quantity of nitre, by which the gold will be preserved at its due degree of oxidation.

It is not essential that gold used for this preparation should be absolutely pure or unalloyed; since neither copper nor silver, when present in small quantities, appears to alter or diminish its colouring power.