58 LETTERS ON NATUEAL MAGIC. 



illuminated in a very high degree, the images 

 which they formed must have been very faint and 

 unsatisfactory. The silver mirrors, therefore, 

 which were universally used, and which are 

 superior to those made of any other metal, are 

 likely to have been most generally employed by 

 the ancient magicians. They were made to give 

 multiplied and inverted images of objects, that 

 is, they were plane, polygonal or many-sided, and 

 concave. There is one property, however, men- 

 tioned by Aulus Gellius, which has given un- 

 necessary perplexity to commentators. He states 

 that there were specula, which, when put in a 

 particular place, gave no images of objects, but 

 when carried to another place, recovered their 

 property of reflection.* M. Salverte is of opinion 

 that, in quoting Varro, Aulus Gellius was not 

 sufficiently acquainted with the subject, and erred 

 in supposing that the phenomenon depended on 

 the place instead of the position of the mirror ; 

 but this criticism is obviously made with the view 

 of supporting an opinion of his own that the 

 property in question may be analogous to the 

 phenomenon of polarised light, which, at a 

 certain angle, refuses to suffer reflexion from 

 particular bodies. If this idea has any founda- 

 tion, the mirror must have been of glass or some 

 other body not metallic, or, to speak more 

 correctly, there must have been two such mirrors, 

 so nicely adjusted not only to one another, but 

 to the light incident upon each, that the effect 

 could not possibly be produced but by a philoso- 



* Ut speculum in loco certo positum nihil imaginet ; 

 aliorsum translatum facial imagines. Aul. Gel. Noct. 

 Attic., lib. xyi., cap. 18. 



