564 Prof. A. A. Michelson on Metallic 



of the metallic colours in birds and insects, is entirely con- 

 clusive. 



It is clear that in all of these carves the descriptive colour 

 corresponds in general to that colour for which the full 

 curve is least steep, and for which the dotted curve is 

 highest ; and is complementary to the colour for which the 

 full curve is steepest and the dotted curve is lowest, as we 

 should expect ; since the former corresponds to high re- 

 flective power, while the latter is characteristic of transparent 

 substances with but moderate reflecting power. 



Exceptions. 

 Morpho alga. 



The measurement of the phase- difference in the light 

 reflected from the blue-winged butterfly {Morpho alga), 

 instead of being zero at normal incidence, had values which 

 ranged from 0*15 to — 0\15, and which were found to vary 

 with the orientation of the specimen. There were also cor- 

 responding changes in the general character of the phase 

 and amplitude curves, all of which showed clearly that the 

 whole phenomenon is considerably complicated by a structure 

 of the scales. 



An examination under the microscope revealed the presence 

 of exceedingly fine hairs (which can only be seen in reflected 

 light) arranged without much regularity with their length 

 parallel with that of the scale*. 



It was at first natural to attribute the blue colour to the 

 light diffracted from these hairs ; and it is not impossible 

 that some of the silky sheen which these butterflies exhibit, 

 is at least in part due to these hairs, whose diameter is much 

 less than a light-wave, and which are therefore in the same 

 relation to the light-waves as the small particles which cause 

 the blue colour of the sky. But the changes in colour with 

 varying incidence, so characteristic of true " surface colours/'' 

 were precisely the same in this specimen, and were practically 

 independent of the orientation ; whereas the changes with 

 the angle of incidence, which should result on the hypothesis 

 that the colour is due to diffraction, should follow an entirely 

 different law 7 . 



Another species of butterfly {Papilio Ulysses) was also 

 examined and found to yield normal surface-colour curves, as 



* There are three varieties of scales, of different shapes. These are 

 arranged in overlapping layers, the outer layer being quite transparent 

 and the lower one opaque. The middle layer is the one showing blue 

 by reflexion and brownish-yellow in transmitted light. 



