162 THE CANADIAN ENTOMOLOGIST 



of scales, it will be found that there is no corresponding change of 

 colour, thus showing that the colour is not due to interference of 

 the light reflected from the surfaces in question. 



On the other hand, an examination by means of the micro- 

 scope reveals the fact that the light and dark scales are not mixed, 

 as are the yellow and black scales of the Euchloids, and hence the 

 blue colour is not due to the mixing of two other tints. The scales 

 are arranged in two layers, the top one consisting of the whitish 

 transparent scales, and the lower one of dark opaque scales of a 

 dark-brown or black. In male specimens there are also a vast 

 number of small sexual scales known as "androconiae" or "plumules," 

 but these do not contribute towards the coloration, and need not be 

 considered here. There are left for consideration three possible ex- 

 planations of the blue effect produced by this arrangement of 

 the scales. 



1. — By diffraction of light from the finely striated upper 

 surface of the top layer of scales in the same way that colours are 

 produced when white light is reflected from the finely ruled mirrors 

 of a diffraction spectroscope. This theory is untenable, however, 

 since the colour would have to range from red to violet, according 

 to the angle of view, and at any one point would only give rise to 

 such colours in a line at right angles to the striations and not in 

 one parallel to them. The effect of diffraction is, however, seen 

 in those species of Lepidoptera which display "sheen." This not 

 only varies according to the angle, but at any one point becomes 

 totalh- invisible when looked at from a point in line with the 

 striations. This sheen, however, is an additional source of colour, 

 and is quite independent of the general tint of the insect. 



2. — ^As an optical illusion due to seeing the dark layer of scales 

 through a whitish transparent upper layer. This explanation will 

 be considered together with the 3rd and last, namely, that it is 

 due to inte-Val colouring matter in the upper scales. It has been 

 held thd4:'v'his latter explanation cannot be correct, because when 

 loosely scattered scales from the blue surface are examined by means 

 of a microscope with transmitted light, the blue effect totally 

 disappears and we see only an equal number of dark brown or 

 blackish scales and yellowish transparent ones, which are some- 

 times nearly colourless. This, however, is not a valid objection, 

 since many colouring matters appear of a totally different colour 



