HISTORY OF GLASS-MAKING.

XI

blue, yellow, green, and orange. There are many shades of the former as well as ofthe latter, particularly of transparent blue, and of opaque blue, yellow, and green.Of opaque colours many varieties appear to be due to the mixture of one colour withanother. In any large collection of fragments, it would be easy to find eight or tenvarieties of opaque blue, ranging from lapis lazuli to turquoise or to lavender, andsix or seven of opaque green. Of red the varieties are fewer; the finest is a crimsonred of very beautiful shade, and there are various gradations from this to a dullbrick red. One variety forms the ground of a very good imitation of porphyry, andthere is a dull semi-transparent red which, when light is passed through it,appears to be of a dull green hue.

The analyses of antique glass of this period, given by Yon Minutoli (p. 31),which were made by Klaproth, show the following results:—-

Opaque Bed from the Villa of Tiberius at Capri.

In 200 grains, Silica .

142 grains.

Oxide of Lead

28 „

Do. Copper .

15 „

Do. Iron

5 ,,

Alumina

2 „

Lime

3 „

195 „

Opaque Green.

In 200 grains, Silica .

130 grains.

Oxide of Copper ,

20 „

Do. Lead

15 „ .

Do. Iron

7 „

Lime

13 „

Alumina

11 „

196 „

Partially transparent Blue

from Capri.

In 200 grains, Silica .

163 grains..

Oxide of Iron

19 „

Alumina

3 „

Oxide of Copper .

1 „

Lime

0.5 „

186.5 „

Another specimen of blue Roman glass, analysed by Professor John, shewedthe presence of cobalt.

It has been ascertained that the rose-coloured glass owes its colour to gold,the violet to manganese, the white and orange to oxides of tin and of arsenic.A specimen of fine opaque red lately analysed in London proves to be colouredby iron.

With these colours the Roman vitrarius worked, blending them in almost everyconceivable combination, sometimes, it must be owned, with a rather gaudy andinharmonious effect.

These combinations of colour were effected in two ways: first, by glasses of