世有透光鉴，鉴背有铭文，凡二十字，字极古，莫能读。 以鉴承日光，则背文及二十字，皆透在屋壁上，了了分明。人有原其理，以为铸时薄处先冷，唯背文上差厚，后冷而铜缩多。文虽在背，而鉴面隐然有迹，所以于光中现。予观之，理诚如 是。然予家有三鉴，又见他家所藏，皆是一样，文画铭字无纤异者，形制甚古。唯此一样光透，其他鉴虽至薄者皆莫能透。 意古人别自有术。
Mengxi Bitan Chapter 19 Qiyong 12
I once saw a see-through mirror. Altogether twenty ancient Chinese characters were carved on its back, and they were undecipherable. If the mirror was put under the sunlight, the patterns and the characters would be clearly reflected on the wall. There were people who wanted to find out the reason for such a phenomenon. In their eyes, when the mirror was being cast its thinner parts cooled down first. As the parts carved with patterns and characters were thicker, they cooled down a bit more slowly, which caused the bronze to undergo deeper contraction. Though the patterns and characters were on the back of the mirror, their traces were still visible on the surface. That was why these patterns and characters were transparent in the sunlight. I once examined this mirror and thought that their explanation was reasonable. I have three mirrors in my own house, which, together with ancient mirrors that I find in other peoples’ houses, are of the same style, the same patterns and characters. However the sunlight can only make one of such mirrors transparent. Other mirrors are also thin, but the light cannot pass through them. I guess that the ancient artisans must have used a special method to make see-through mirrors.
世上有能透光的镜，镜背面有铭文，共 20 字，字体非常古奥，没有人能读懂。用这面镜子承受 日光，那么背面的花纹和 20 个字都透射到墙上，清清楚楚。有人探求它的道理，认为是因为铸造时薄 的地方先冷，唯独有花纹和字的地方较厚，冷得慢，以致铜收缩要多一些。铭文和花纹虽然在背面， 但镜面也会隐约可见。所以镜子放在日光中就会现出来。我观察这面镜子，道理确实如此。但是我家 有三面镜子，另外看见别人家收藏的镜子，都是一样的，图案铭文没有一点差异，样式都非常古老。 只有那一面镜子能透光，其它的镜子也有非常薄的，却都不能透光。由此想来，古人在造这面镜子时自有别的方法。