THE MEDIEVAL CATHEDRALS 427 rjup on the roof were carefully sculptured in every muscle and fold of their skins, behind as well as in front and below las well as above. Either the artist thought that God would see it even if men did not, or he executed the work so thor- oughly because he liked to do it. Such were the sculptors I who in carving floral designs about a capital would amuse | themselves by occasionally converting a petal into a face or hide an imp in a mass of foliage. In the choir stalls the under sides of the folding seats were sometimes covered with the most exquisite wood carving. In place of the Byzantine mosaics the Gothic cathedrals jhad transparent colored designs in their stained-glass win- dows. We first learn of the making of stained glass stained I from a treatise on various industrial and artistic glass processes written by the monk Theophilus in the twelfth century. In the thirteenth century the coloring matter was diffused into the mass of glass while it was yet in the melting- pot, so that it was colored all through and was of a brighter hue than in later times when it was merely tinted upon the ! surface. The windows were made up of small bits of glass which were pieced together and held in place by leads. This thirteenth-century glass was imperfect in character, and since the fragments of it differed further in shape and size, the rays of light in passing through them were broken up the more, and there was much blending of the different colors and very brilliant effects were produced like the glittering of jewels. The leads were skillfully employed to form the out- lines of the human and other figures depicted in the design, whereas later, in the sixteenth century, when large plates of painted glass were used, the leads were arranged in mechani- cal squares and would sometimes run across a saint's face or sever his body. In the fourteenth century it was discovered how to stain glass yellow by means of silver; before this, purple had been the favorite color, but it did not admit as much light. It was also discovered early in the fourteenth century that, by dipping the blow-pipe first into liquid glass of one color and then into that of another color, a sheet of glass could be blown of one color on one side and another
