164 THE CANADIAN ENTOMOLOGIST. 



the hypothesis that the colour is due to pigment or dye, but is 

 not explainable if the colour were due to the superposition of white 

 transparent scales on a background of black or dark-brown ones. 

 Finally, further light may perhaps be thrown on the subject by 

 actually analysing the colours as they are emitted from the wing, 

 by means of the spectroscope. To do this I improvised a small 

 spectroscope attachment to an old and low-power microscope in 

 my possession. When I examined the wings of white butterflies 

 or moths, the whole spectrum was clearly visible. When I ex- 

 amined the red portions of such butterflies as Vanessa atalanta, 

 Parnassius apollo, the under wing of a Catocala or the red spots 

 on a Zygaenid, the red portion of the band remained brilliant, 

 but the yellow and green was greatly weakened and the rest was 

 practically invisible. Orange and reddish-brown butterflies, such 

 as the Argynnids or Chrysophanus hypophloeas, gave strong red 

 and orange bands, weaker yellow and the rest of the spectrum 

 was very much weakened, but nevertheless contributed something 

 to the total efifect. When I examined the wings of L. pseudargiolus, 

 I found the green, blue and violet strong, as I expected, but there 

 was some red and yellow present also, but weak. The intensely 

 blue wing of L. hellargus, however, gave brilliant green and blue 

 bands, rather weaker violet and very weak red-yellow and orange. 

 When it is considered that these experiments were made with 

 the light of an incandescent gas burner, the results are not sur- 

 prising. Such light is known to be deficient in the blue and violet 

 rays, but are strong in the yellow and red rays, although ,of course, 

 the entire spectrum is clearly visible. Again, the presence of red 

 in every case is, I think, partly due to total reflection from the 

 metallic surfaces of the scales at angular points and from the glass 

 itself which covers the wing in those cases where a slide was made 

 for the purpose of examination. Also it is notoriously difficult to 

 produce artificially a green or blue colour free from red, though it 

 is easy to produce red free from blue and green, and this difficulty 

 may also be felt by nature in preparing the natural tints of insects. 

 In any case the strongly-marked blue end of the spectrum and 

 the much-weakened red end show clearly that the scales them- 

 selves are inherently blue, in that the total colour emitted by the 

 wing is actually blue and is not a mixture of rays from a brown 

 or black surface seen through a white transparent layer, which 



