"The fact of coloration," says Mr. Gaffield, "or change of color or shade by sunlight being established, must we not transfer some of our praise for the old artists in glass to the wonderful pencil of the brightest luminary of the heavens, which, during the centuries, has noiselessly but unceasingly been at work, deepening and mellowing the colors of all the windows of the venerable cathedrals of the world? We do not see to-day the glass as it was when it came from the artist's studio or the glass-factory. The dust of ages has accumulated upon its surface, the corroding tooth of Time has eaten into it; but how often has the wonderful alchemy of the sunlight done more in penetrating beneath the surface, and changing the shade or color of the entire body of the glass!"
The cause of the interesting phenomena above stated has not yet been explained, although not a few theories have been advanced to account for it. The change in color has been variously attributed to the presence of oxide of iron, to arsenic, to sulphur, and to oxide of manganese, in the constituent materials of the glass. But the true solution remains yet to be given. That the effect is not due to heat, but solely to the actinic rays of the sun, is shown by the fact that no change of color is produced in the glass when exposed to heat; while, on the contrary, after the discoloration has been produced by solar light, the colors thus acquired disappear under the action of heat, and the glass assumes its normal color. This process may be repeated indefinitely, the change of color being produced by solar light, and the original color restored by heat. It has also been shown that the effect is not produced by air or moisture.
Prominent among other interesting facts shown by these experiments is the varying effect of sunlight during each season, and each month of the year. The comparative actinic power of the rays during each month is shown, at the end of the year, by the comparative depth of yellow or purple color produced. The actinic effect increases from January to July, and decreases after the latter month; the greatest effect is observed in the summer, and the least in winter; in the spring and autumn it is about equal, being midway between that of winter and summer.
The comparative power of different kinds of glass to transmit the actinic rays was also tested. Of the colored glasses it was found that blue transmitted the most, purple less, and red and orange the least; which shows the propriety of the preference given by photographers to blue glass for skylights, because it transmits the blue rays, which exert the most active power. Mr. Gaffield's observations have not been confined to the glasses exposed on his own roof and window-sills; but he has been earnest in collecting, from various sources, specimens of window-glass that have been exposed for greatly-varying periods. By observing the portion of the pane exposed and that protected by the sash and putty, a comparison is afforded between the original and
