Bertin, who wrote two years later, confirmed the conclusions arrived at, and repeated on other mirrors the experiments of Ayrton and Perry, as well as those of Person and Govi. Finding that the distortions produced by heating were particularly efficacious in bringing out the magic qualities, he sought to imitate them mechanically, and, with the aid of the optician Duboscq, constructed an apparatus for conferring temporary convexity upon mirrors by mounting them against an air-tight back, with a cavity behind into which air could be driven by a force-pump. With this apparatus he examined the effects that are produced by drilling cavities, cutting grooves, and etching depressions in the backs of mirrors. Laurent, taking up this line of suggestions, produced magic mirrors of thin glass, etched into patterns at the back, and silvered on the front face. These, when mounted on an air-tight back, could be made to show either positive or negative figures by diminishing or increasing the pressure of the air behind the mirror by means of a simple india-rubber pear attached by a flexible tube to an aperture in the back; the pressure of the hand sufficing to bring out the optical effects. Laurent further showed that by taking an ordinary piece of mirror glass (patent plate glass silvered at the back) magic effects could be produced by lightly pressing against the back hot pieces of metal on which a raised pattern had been engraved. In this case the portions of the mirror nearest in contact with the hot metal became hotter than the other parts, and expanding more, set up minute differences of curvature sufficient to concentrate the reflected rays of light from the heated parts.