228 COLOUR-PllODUCTION IN THE FEATHERS OF BIRDS. 



fluid (alcohol, xylol, chloroform, oil, etc.), but strong acids in 

 time cause the colours to change towards the red end of the 

 spectrum. 



The most penetrative fluid which I have tried is the solution 

 of iodide of mercury in iodide of potassium. This, when concen- 

 trated, rapidly destroys the feather substance, but in dilute 

 solution merely penetrates into the interior. The gorget-feathers 

 of the Humming-Bird (from Costa Rica) reflect a brilliant lilac, 

 i. e. a mixture of red and blue, but after a few hours' immersion 

 in the iodide solution, the red disappears and the blue changes to 

 a very bright green. 



All these metallic colours shift towards the blue as the angle 

 of incidence of the light increases, as do the ordinary colours of 

 thin plates ; but this is not a proof that both have the same origin, 

 for many of the aniline colours when in thin dry films show 

 somewhat similar changes depending on the angle of incidence. 



Michelson in America has compared the metallic colour of 

 some beetles with those of the anilines, and has given reasons 

 (connected with similarity of the polarization of light reflected by 

 both) for believing that the origin of the colours in the two cases 

 is of the same kind. 



The pressure test, however, seems to make this conclusion 

 invalid. 



The colours seem to me to be more allied to those of Lipmann 

 films, in which layers of reduced silver are spaced at half wave 

 intervals, and in the case of metallic feathers I believe that one 

 or two layers of optically dense material are the sources of 

 interference. 



Although half wave-lengths can be readily resolved by high- 

 power microscopic objectives, it is almost impossible to cut sections 

 thin enough (viz. less than -00002 in.) to use with such powers. 

 At least I have cut many hundred sections, but although in some 

 cases a laminated structure seemed to be present, this was due to 

 a diffraction effect, as was evident from the changes in the dimen- 

 sions of position of the apparent lamina which occurred with the 

 change of focal adjustment. 



In the case of the Lipmann films, the layers of reduced silver 

 are readily seen if the sections are expanded by wetting, though 

 I have not been able to resolve them satisfactorily when dry. 



