He was able also to reproduce writing in luminous lines on a screen or wall by the following device. A piece of thin sheet lead (tea-chest lead) was laid on a bed of blotting-paper; and upon this was written with a common lead pencil any word desired, which therefore was slightly indented into the lead. The sheet was then pressed lightly against an ordinary piece of mirror glass and heated from behind by pressing against it a disk of hot metal. The written letters touching the back of the mirror warmed it, and made it curve at these parts. Hence, placed in an appropriate divergent beam of light, it threw the handwriting upon the wall. Following the hint given by Bertin’s research upon the effect of bending mirrors by air-pressure at the back, the author finds that a similar effect is even more powerfully produced by simple mechanical pressure. He took a mirror which, though it had an excellent reflecting face and a well-raised pattern on the back, showed no magic properties, and having clamped it in a wooden frame, he applied screw pressure behind to force against the back a slightly convex piece of soft wood covered with a pad of cloth. On turning the screw the mirror at once became magic, and was found, even after the screw had been released, to retain some part of its magic property. It was again distorted by screw pressure, and while violently bent was heated to anneal it somewhat. On removing the pressure it was found to retain permanently all the qualities of a good magic mirror, though it is slightly more convex than mirrors ordinarily are. Since then it has been found that many mirrors which, when purchased, showed no magic properties, can be converted into magic mirrors, some by application of screw pressure, some by mere bending by hand over the knee, some by burnishing under pressure the pattern at the back. Engineers are so familiar with the circumstance that when metal castings are rapidly cooled the internal parts are in a state of strain, that they are not astonished to find that after a casting has been surfaced accurately on one face the figure may alter slightly by the mere release of internal strains in the slow annealing of time. It is quite possible that this too occurs with Japanese mirrors, and that some of them may acquire with the mere lapse of time magic qualities that they do not show at first when newly polished.