34 LETTEES ON NATURAL MAGFC. 



as if it had been formed by reflection from a piece 

 of mirror glass, though of course much less bril- 

 liant, and the position of the image proved that 

 it must be formed by reflection from a perfectly 

 flat and highly polished surface. But where such 

 a surface could be placed, and how, even if it were 

 fixed, it could reflect the image of the candle up 

 through my head, were difficulties not a little per- 

 plexing. Thinking that it might be something 

 lodged in the eyebrow, I covered it up from the 

 light, but the image still retained its place. I 

 then examined the eyelashes with as little success, 

 and was driven to the extreme supposition that a 

 crystallization was taking place in some part of 

 the aqueous humour of the eye, and that the 

 image was formed by the reflection of the light of 

 the candle from one of the crystalline faces. In 

 this state of uncertainty, and, I may add, of 

 anxiety, for this last supposition was by no means 

 an agreeable one, I set myself down to examine 

 the phenomenon experimentally. I found that 

 the image varied its place by the motion of the 

 head and of the eyeball, which proved that it was 

 either attached to the eyeball or occupied a place 

 where it was affected by that motion. Upon 

 inclining the candle at different angles, the image 

 suffered corresponding variations of position. In 

 order to determine the exact place of the reflecting 

 substance, I now took an opaque circular body 

 and held it between the eye and the candle till it 

 eclipsed the mysterious image. By bringing the 

 body nearer and nearer the eyeball till its shadow 

 became sufficiently distinct to be seen, it was 

 easy to determine the locality of the reflector, 

 because the shadow of the opaque body must fall 



