of some Brilliant Animal Colours. 109 



underlying pigment maybe removed without much affecting 

 the dye, or the structure, which is the seat of the coloration. 



Several experiments were made to test whether air-cavities 

 existed. For this purpose the wing-case was exposed for 

 some time to the action of a vacuum, into which afterwards 

 water or benzol was admitted. But no distinct evidence of 

 the penetration of liquid could bo recorded. I understand 

 that Prof. Poulton has had a similar experience. 



Again, it has been noticed by Mr. H. Onslow and myself 

 that considerable pressure fails to alter the colour of beetles 

 and of the wings of some iridescent dragon-flies, though 

 (Poulton, Mallock) effective in some other cases. It 

 would seem that the hypothesis of air-cavities must be 

 abandoned. 



In the absence of air-cavities the alternating- structure 

 demanded by the interference theory would require two 

 kinds of matter capable of resisting pressure and of sensibly 

 different refractive indices. Probably both would be solids; 

 and since the range of relative index is then much restricted, 

 the brightness of the reflected light could hardly be explained 

 without supposing more than the two or three alternations 

 which might suffice were air in question. Mr. Onslow thinks 

 that there may then be a difficulty in finding room for the 

 alternating structure and the protective covering. 



An important question is whether the change of colour 

 with angle is such as can plausibly be attributed to a periodic 

 structure. As Walter points out, a good deal depends upon 

 whether, or not, there is a limitation upon the obliquity of the 

 rays within the thin plate, or plates. In the ordinary arrange- 

 ment for Newton's rings there is no limitation, the direction 

 in the air- film being parallel to that of the rays before 

 incidence upon the first plate. The optical retardation may 

 then vary from its maximum at perpendicular incidence to 

 zero at 90° obliquity.' According to this, it should always 

 be possible to push the colour out of the spectrum at the blue 

 end by sufficiently increasing the obliquity, but it must be 

 remembered that unless special provision is made the colour 

 effects would be overlaid by the white light reflected at these 

 angles from the first glass surface encountered. 



From what we have seen in the case of the beetle colours 

 where we must suppose that the refractive index does not 

 differ greatly from that (1*6) of the chitin, there is a limit to 

 the obliquity within the thin films even when externally the 

 incidence is grazing. If 6 be the angle in the thin film and 

 p the refractive index, the retardation is proportional to 

 cos 0, and in the limiting case 



COS0 = v/(/X 2 -l)/ya. 



