The glass itself is in all cases remarkable for the number of minute bubbles contained in it; in some of the lamps these bubbles are so numerous that the material is barely to be classed as transparent. In colour the glass varies from a pronounced bottle-green to an amber tint; it is more rarely of a greyish white. The size of many of the lamps and bowls must have necessitated the use of large melting-pots as well as considerable skill in blowing and manipulation. The irregular form so often to be observed in both lamps and vases is more likely to be the result of a partial collapse during the melting on of the enamels, than of any defect in the original piece as it came from the glass-blower’s hands. In composition, to judge from the analysis of a fragment of a Cairene lamp made by Dr. Linke of Vienna (Schmoranz, p. 42), this Saracenic glass is essentially a normal soda-lime silicate with 69 per cent. of silica, 15·4 per cent. of alkali, and 8·6 per cent. of lime, thus far resembling the ordinary Roman type. The specimen examined, however, contained in addition to the lime as much as 4 per cent. of magnesia. As Dr. Linke points out, the presence of this last base would hinder the complete fluidity of the glass in the pots and make it difficult to get rid of the bubbles. But whether the presence of this earth in a single specimen is in itself sufficient to prove the non-Egyptian origin of these lamps as a class is another question. The fact that nearly one per cent. of manganese was found in this glass is of interest, as it shows that some attempt had been made to ‘cleanse’ the metal. As regards the enamels on this Saracenic glass, we find that, with one important exception, they resemble generally in composition and character those employed at a later date by the Chinese in the decoration of their porcelain[109]—we have a readily fusible flux containing much lead coloured by various metallic oxides. The opaque red is given by oxide of iron, the green by oxide of copper, and the yellow by antimonic acid. The presence of this last substance is of interest: Dr. Percy found antimony in the glaze of Assyrian bricks, and I have taken for granted that it is the source of the yellow in the primitive glass of Egypt. The opaque colours, including the white, are probably produced by the addition of a little oxide of tin to the flux; Dr. Linke, however, does not seem to have found that metal in his analysis.