OPTICAL PHENOMENA OF ANCIENT DECOMPOSED GLASS. 199 



of the two pencils are those of thin plates, and, as might have been expected, are 

 complementary. Each of the separate films has a colour of its own depending on 

 its thickness ; and when they are numerous, the resulting tint is of a very compo- 

 site character.* The transmitted tints are brightest when the light is incident 

 perpendicularly. They vary at oblique incidences, and disappear wholly when 

 the plate is considerably inclined. Although the elementary films adhere with 

 such force that it is difficult to separate them , they are not, as I have already 

 stated, in optical contact. The plate of air which lies between them exhibits in a 

 very beautiful manner numerous fringes depending on its thickness, as shown in 

 Plate XI. fig. 12. When we apply a drop of water, or alcohol, or oil, to one of the 

 compound films, the fluid advances irregularly between the elementary films, form- 

 ing a prismatic line of various colours as it proceeds. When the film has been per- 

 meated by the fluid, its colours entirely disappear. When the water or the 

 alcohol is evaporated by the application of heat, or dry air at one edge of the 

 wetted film, the colours reappear, and the moving prismatic line separates the dry 

 from the wet portion of the film. The origin of the prismatic fringe is easily ex- 

 plained. If the fluid enters or quits the space between two of the films more 

 quickly than it does between other two, the colour at that place must differ from 

 that of the compound film. The line of the advancing or retiring fluid must 

 therefore consist of several colours depending on the number of elementary films, 

 whose actions are not affected by the fluid. 



The following are the colours of the prismatic line, or the succession of changes 

 from the colourless transparency produced by the complete absorption of the 

 alcohol to the original tint of the film : — 



Original Tint. Succession of Changes, &c. 



Red. Pale green, green, blue, pink, dark red. 



Rich orange red. Pale blue, blue, pink, pink nearly opaque. 



Yellowish orange. Greenish, bluish, pink, orange red, deep orange red, paler orange red. 



Bright yellow. Pale pink, bright pink, orange yellow, orange, orange yellow. 



Yellowish gi-een. Pale yellow, dirty yellow, deep green. 



Pale bluish green. Bright green, growing paler and paler. 



Deep blue. Dirty yellow, dirty green, blue, bluish green, blue. 



Colourless. Pale green, brighter green, pale green. 



If we use Canada Balsam, or Balsam of Capivi, we sometimes see two waves 

 or prismatic lines, the absorption of the fluid taking place more rapidly between 

 one portion of the elementary films than it does between another portion of the 

 same films. If we use so little balsam that it advances only through half the 

 length of the film, we have the half (A) colourless, and the other half (B) of its ori- 

 ginal brilliant colour ; and if the balsam indurates, the wave of colour which sepa- 

 rates the two portions will be permanent. If we drive out some of the balsam 

 from (A) by alcohol, it is curious to see under the microscope the particles of the 



* See Phil. Trans., 1837, p. 249. 



