Up: Glassmaking Reminiscences 82 of 123 ·i ·23 ·48 ·73 ·98 ·iii ·24 ·49 §74 ·99 ·iv ·25 ·50 ·75 ·100 §1 ·26 ·51 §76 ·101 ·2 ·27 §52 ·77 ·102 ·3 ·28 ·53 §78 ·103 ·4 ·29 ·54 ·79 ·104 ·5 ·30 ·55 §80 §105 ·6 ·31 ·56 ·81 ·106 ·7 ·32 ·57 ·82 ·107 §8 §33 ·58 ·83 ·108 ·9 ·34 ·59 ·84 ·109 ·10 ·35 ·60 ·85 §110 ·11 §36 ·61 ·86 §111 ·12 ·37 ·62 ·87 §112 ·13 ·38 ·63 ·88 ·113 ·14 ·39 ·64 ·89 §114 ·15 ·40 ·65 ·90 §115 ·16 ·41 ·66 ·91 ·116 ·17 ·42 ·67 ·92 §117 ·18 §43 ·68 §93 ·119 ·19 ·44 ·69 ·94 ·121 ·20 ·45 ·70 ·95 ·123 ·21 ·46 ·71 §96 ·22 ·47 ·72 ·97

speaking of this superiority, attributes it to what moderns regard as defects. In the first place, much of the ancient glass is of unequal thickness, and so presents convex and concave parts, which refract the light differently and produce an agreeable effect. In the next place the old colored glass is not a colorless glass, to which has been added the particular coloring material, such as protoxide of cobalt, &c. Old glass contains a good deal of oxide of iron, which colors it green, and to this must be attributed the peculiar effects of antique glass, colored by cobalt and manganese. M. Chevreul appears to think that modern stained glass is too transparent to produce the best effects. M. Regnault, the chemist, has recommended that all this kind of stained glass should be cast, to avoid the monotonous effect of plain surfaces on the light; and also that foreign substances should be mixed with the glass to diminish its transparency.     Many attempts have been made to color with ruby or other colors gas shades so as to throw on surrounding objects the color of the glass; but in no case has the ray of light passing through colored glass, to refract the shade, been successful.     But when a ray of solar light is passed