288 Mr. Ainger on the 



make this experiment without the intervention of glass, as it 

 is not easy to obtain either a sphere or cylinder of water un- 

 supported by a containing vessel, an equivalent form may be 

 procured by placing a drop between two small surfaces, as in 

 Fig. 5., near the middle of which the tangents of the opposite 



Fig. 5. 



surfaces will be parallel, which is, of course, all that is required. 

 I used the ends of two black lead pencils for this purpose ; but, 

 finding that a thin glass bulb did not sensibly affect the 

 results, I made the observations with that as being more con- 

 venient and manageable. In this way it will be seen, that, 

 although the rays transmitted near the maximum and minimum 

 angles of deviation are more brilliant than the others, yet the 

 difference is one only of degree, and is not sufficient to render 

 it probable that the latter are made invisible by mere diverg- 

 ence. The ray, after one internal reflexion, becomes dis- 

 tinctly visible as soon as it ceases to be confounded with the 

 ray obtained by reflexion from the first surface, as in Fig. 6. ; 



Fig. 6. 



and it continues to increase in brightness, but without sensible 

 colour, till the bulb arrives at the maximum angle of devia- 

 tion, for the violet rays of the primary bow, as in Fig. 7. ; 

 after which the light becomes more feeble and coloured, till it 

 vanishes altogether in a faint red. The same sort of progres- 

 sion is observed by commencing with a position of the bulb 

 directly between the eye and the light, in which there will be a 



