106 On the Optical Illusion of the 



tion by the inverting eye-piece, as in Fig. 15. Then, if the 

 cavity is sufficiently deep, and if we place the point of our 

 finger in the cavity, the evidence which this gives us of its 

 being a depression, is superior to the evidence of its being a 

 cavity arising from the inversion of the shadow ; the apparent 

 elevation will of course sink into a depression ; but the mo- 

 ment the finger is withdrawn, it will again rise into an eleva- 

 tion. If the cavity is a long groove, the part not touched 

 by the finger will appear elevated, while the part touched by 

 it will appear depressed ! 



Having thus considered some of the principal phenomena 

 arising from the inversion of the object, we shall now proceed 

 to explain some analogous facts which are owing to the semi- 

 transparency of the body. If M N, Fig. 21, is a plate of 

 mother-of-pearl, and A a cavity ground or turned in it ; then 

 if this cavity is illuminated by a candle C, or by a window at 

 C, in place of there being a shadow at the side s, as there 

 would have been had the body been opaque, there is a quan- 

 tity of refracted light seen along the whole side s, next the 

 candle. The consequence of this is, that the cavity appears 

 as an elevation when seen only by the naked eye, as it is only 

 an elevated surface that could have the side s illuminated. 

 The fact which we have now stated, is, we think, a very im- 

 portant one, in so far as it may affect the labours of the 

 sculptor. In some kinds of marble, the transparency is so 

 great, that the depressions and elevations in the human face 

 cannot be represented by it with any degree of accuracy ; and, 

 consequently, transparent marble ought never to be used for 

 works of any importance. 



Illusions arising from the same cause may be observed 

 even when the surface of the object is perfectly plain and 

 smooth. If M N, Fig. 22, is the surface of a mahogany 

 table, MNnma section of it, and abc a section of a knot in 

 the wood, then it often happens, from the transparency of the 

 thin edge at a, next the candle, that that side is illuminated 

 while the opposite side at c is dark, the eye being placed in 

 the plane of the section abc. The consequence of this is, 

 that the spot abc appears to be a hollow in the table. 



Hence arises the appearance in certain plates of agate, 



