September 30, 1920] 



NATURE 



151 



plates themselves, so that they also are yellow. 

 The addition of this pigment converts the blue into 

 a green, and the effect of suppressing it is seen in 

 O. urvilliana, a powdery blue insect, a section of 

 which, 4, is shown in Fig. 2. It is pale blue on the 

 upper surface, probably because the chitin plates 

 are broad, but underneath, where the plates are 

 like those in section 5, and contain a little yellow 

 pigment, it is pale green. In some other insects 



Papilio ulysses. — The insects the colour of which 

 changes towards the violet at grazing incidence do 

 not have a periodic structure such as that de- 

 scribed ; but they invariably exhibit a somewhat 

 thick superficial layer of clear chitin. The scales 

 of many iridescent Papilios belong to this group, 

 as, for example, the satin-blue scales of P. ulysses, 

 section ja, Fig. 2. The layer of chitin, b, is clearly 

 too thick to cause colour, and no finer structure 

 could be made out, even with an objective capable 



Fig. i—Omiktfttru faradijia $. Body and hind-wingt gold 

 for<-wia(< gron and black. (^ naiatal liic) 



the pigment is granular, as at 5, in section 6, being 

 situated in the body of the scale, but not in the 

 chitin plates. This is the case with those Pierids { 

 •uhich have magenta tips to their fore-wings, as 

 well as with the beautiful purple-eyed Callitaera 

 esmeralda, which has scaleless but iridescent fore- ! 

 wings (Fig. 5). 



•v / 



fio. i.—Xffira-iti .iri.ym (ihe Larg* BIm) MiU and fomale, aad •nd« lido 



Nat Mrml %U9. ) 



Hlack is naturally a very important pigment, 

 for it often serves the purpose of absorbing white 

 .light, which otherwise, being reflected, would 

 fgreatly desaturate the colour. Thus the white 

 ' spots in many inserts, such as Morphn cypris, 

 liave exactly the same structure as the deep blue 

 jParts, except for the absence of the absorbent 

 f backing of hl.ick pit^'mcnt in the lower layers of 

 ffthe scalf, 



NO. 2(157. VOL. 106] 



Fig. i.—Ca/li/ae'-, t,mmlla. The win-s ar<! scalclew and faintly 

 indecent. The ayes of the hind-wing« arc mag nta. (Natural 



of separating the rulings of a grating o-2i/i 

 apart. This is less than one-half the mean 

 wave-length in air, which seems to preclude the 

 existence of air-films. Nevertheless, the surface 

 layer might contain three or four half-wave-length 

 plates of chitin, placed exceedingly close together, 

 though not actually in optical contact. It is,' 

 however, doubtful whether in this case pressure 

 experiments and immersion in highlv 

 refractive fluids would produce the effect's 

 observed. 



Very strong confirmation of the 

 existence of a periodic structure, of 

 some description, parallel to the surface, 

 can be obtained by cutting oblique 

 sections of /'. ulysses. The.se sections, 

 3h (Fig. 2), show the ragged edges of 

 three or four layers of chitin, mm, 

 which come into focus successively on 

 lowering the objective. 



Salamis parhassus. — There are .some 

 insects, like the pale pink Salamis 

 parhassus, which have scales that 

 might owe their colour to the thin 

 double membranes of which they are 

 composed, a and b, section 7 (Fig. 2). If 

 these single films of chitin really cause 

 the colour, it is diflScult to account for 

 the uniformity of shade, which ought to vary with 

 the inevitable differences of thickness inherent in 

 an organic film. Such variegated colours are 

 actually found in the lower membrane of many 

 black scales, such as those of Vrania fulgens, 

 h, section to (Fig. 2). 



There are, however, other insects, such as 

 many Lycsenids (Figs. 4 and 6), the iridescent 

 scales of which have mcmbr.ines too thick to pro- 



of lama. 



