THE ACTION OF TWO SUPERPOSED CRYSTALLOID BODIES. 343 



proximately the same, though the field of view becomes more 

 and more dark, until at length with an angle of 90 complete 

 darkness, or at any rate when the two bodies are not exactly 

 equal the greatest darkness, is found. The passage from one 

 tint to the other certainly takes place in a somewhat different 

 manner to that in Newton's Table, generally as if the two colours 

 were mixed in ever- vary ing proportions ; in the majority of cases, 

 however, the rise or fall cannot be mistaken. 



If, for instance, we place two selenite plates, each of which 

 yields light blue of the first order, together in the given way, we 

 obtain f or e =0 yellow I., for e = 22J light yellow, fore = 45 

 white : finally, for e = 67J bluish white. These colours evidently 

 represent only those tints taken in reversed order to yellow I. 

 A few additional examples are collected together in the following 

 table : 



It is evident from this table that the colour we obtain in the 

 position e = 45 agrees tolerably throughout with that of the single 

 plate, or very approximately. Inasmuch as the acceleration colour 

 for the position e = is to be regarded as given, we can pre- 

 determine the changes which the gradual increase of e produces 

 in any given pair of plates. These changes always consist in a 

 gradual but direct passage of the acceleration colour into the 

 colour of the single plate. 



It now remains for us to combine the action of a firmly bound 

 pair of plates with that of a selenite plate, or in general of a 

 third double-refracting body. We have seen that the pair of 

 plates act, to a certain extent, like a single body whose ellipse of 

 elasticity approaches the more nearly to the circle the greater 

 the angle e. If we suppose that the axes of this ellipse are the 



