150 



NATURE 



[September 30, 1920 



narrow strips of chitin, a, Fig. 2, ic. The chief 

 peculiarity of these films is that they are placed 

 at right angles, and not parallel, to the surface 

 of the scale or wing, as may clearly be seen from 

 the sections. This results in an obvious con- 



la 



U.V-*-*J.A% 



3i> 



•10 



\>i\r<0^ 



fWYiYNfO. 



If E-i^"'^ 



Flo. 



/a, Upper scale of Morpho mette^aui. a, ilrije ; h, lower membrane. 



ib, Lowi-r scale of M. vunelaus, a/, air-tilm ; a, chitin film ; c, pigmented striae ; 

 d, band* of chitin joining upper and lower membrane": ; h, lower membrane. 



IC, Longitudinal section ot the last, a, chitin film seen from the side. 



2a, Central spot of Chlorippe taurentia. a, tall films of chitin. 



2i>, Periphery of the above, showing tx, bosses of chitin. 



3a, Scale of Papilio ulyssts showing a, wave like stria: ; b, transparent layer of 

 chitin ; c, pigmented layer of chitin. 



3f>, Diagonal section of the same. a, stria ; r, pigmented portion of strias ; 

 mm, films of chitin. 



4, Blue scale of Ornitkep era urviiiiana. The chitin is colourless. 



^, Green scale of O. posei-.on. The whole scale, including the chitin films a, 

 contains a yellow pigment. 



6, Magenta scale of Callitaera esmeralda, showing granular pigment j ; and a cleft, 

 or division c, in the films of chitin a. 



7, Section through scale, and, tip of scale of Salamis parkassus. a, upper 

 membrane ; b, lower membrane. 



(f. Metallic golden scale of /?/(7«/yw«tf. _ ^, large air.space. 

 q. Iridescent scale of Lycaena uarus stained with carbol-fuchsin. 

 10, Black under-scale of Urania fulgens. h, transparent, iridescent membrane. 

 //, Partially depigmented scale of Evphta deione. 



All these sections were drawn to t le scale M=t*5 mm. with Zeiss 2 mm. apochromat, 

 N.A. I "4, ani Comp. Oc. 



sequence, which goes far to prove that the plates 

 are the true cause of colour. It has always been 

 said that the colour of iridescent insects changes 

 towards the violet end of the spectrum, like a flat 

 soap film, when the wing is turned from the normal 



NO. 2657. VOL. 106^1 



to grazing incidence. This should be true if the 

 colour were produced by films parallel to the wing 

 surface, but if they were at right angles to it, the 

 reverse should be the case ; that is to say, the 

 colour should approach the red end of the spec- 

 trum as the angle reaches the grac- 

 ing incidence. 



The variation of colour in a num- 

 ber of insects was measured in wave- 

 lengths, and most of them were 

 found to fall into two groups^ — 

 (a) those with the periodic structure 

 just described, which at grazing in- 

 cidence reflected the longest waves, 

 and {h) those without this structure, 

 which reflected the shortest waves 

 when in the same position. Further, 

 it was possible to show that the dis- 

 tance between two consecutive plates 

 of chitin was from o-6/i to lo/i,* and 

 since the chitin plates often show a 

 line of cleavage, c. Fig. 2, section 6, 

 so that they appear double, this space 

 may contain two films of chitin and 

 one of air. Thus the plates may be of 

 the most efficient thickness for pro- 

 ducing colour {i.e. one-half wave- 

 length), which for chitin is o-iy fi, 

 and for air o-25/x, or o-6/i for the 

 three plates. 



Chlorippe laurentia. — The edges 

 of the dull green patches on this in- 

 sect's wings are brown at normal in- 

 cidence, but the whole area becomes 

 a brilliant green at grazing inci- 

 dence. This illustrates the effect of 

 the height of the chitin plates on the 

 quality and saturation of the colour. 

 The central area, which is always 

 green, is shown in section 2a, 

 Fig. 2, where the chitin plates are 

 the usual height. In the area which 

 is brown at normal incidence, sec- 

 tion 2b, Fig. 2, the chitin plates, a, 

 have become merely bosses. If a ray 

 must traverse three or four films to 

 give a certain depth of colour, it is 

 obvious that, in the case of sec- 

 tion 2b, it will do this only at very 

 oblique incidence ; whereas, in sec- 

 tion 2a, this will happen with rays 

 more nearly normal to the surface. 



Ornithoptera poseidon (Fig. i). — 

 The males of this gorgeous species, 

 and of the even more remarkable 

 O. paradisea {Fig. 3), are emerald- 

 green, and they illustrate the effect 

 of combining structural and pig- 

 mentary colours. The narrow plates 

 of chitin, a, seen in section 5 of Fig. 2, 

 produce a blue colour, but the body of the scale 

 is dyed by a bright yellow pigment, the colour of 

 picric acid. This pigment extends into the chitin 



' iju = T^gnmm. The section* in Fig. 2 were all drawn to the scale 

 ^=i"S mm., so that the relative distances maj be seen at a glance. 



V).M»^. 



^' 





r<)-^' 



