l82 



NATURE 



[October 7, 1920 



where it appears, as it were, to have been looped 

 up. The bright green areas then correspond to the 

 thickest portions of the scale. This is evidently 

 only another example of the intensification of the 

 colour produced by an underlying screen of dark 

 pigment, which absorbs the excess of white light 

 that would otherwise be reflected to the eye, 

 causing the colour to become much desaturated. 



Diffraction of Light by "Gratings." 



Colours due to barred structures, or "gratings," 

 which diffract light in the usual way, cannot be 

 said, in Lepidoptera at least, to be of very great 

 importance. They do, however, often produce 

 effects which, though of secondary value, con- 

 tribute a good deal to the final result. The fact 

 that most scales appear to be marked with striae, 

 which form gratings of suitable dimensions, has 

 sometimes given rise to the idea that most insect 

 colours are produced in this way. This is evi- 

 dently not the case, for iridescent scales are 

 sometimes smooth, and, moreover, plain black 

 and white ones are often striated. Impressions 

 or replicas were therefore taken of many scales 

 in a special preparation of collodion, in order to 

 isolate the effect of the surface structure from any 

 other colour-producing factors. Good "gratings," 

 showing normal lateral spectra, were obtained 

 from most insects, such as the Large White (Pieris 

 brassicae), but if the film was dyed, the colours 

 became feeble or disappeared. This indicates that 

 diffraction colours are, as might be expected, dis- 

 cernible only on very pale or colourless scales. 

 The existence of diffraction colours can be clearly 

 demonstrated by the following experiment. An 

 impression was made from the pale blue surface 

 scales of Morpho achilles, in such a way that at 

 least one patch adhered to the collodion film. On 

 tipping the grooved film so as to cause the spec- 

 trum colours to pass across the patch of scales, it 

 could be seen that their normal blue colour became 

 intensified in the violet region of the spectrum, 

 changed to mauve or pink in the red region, and 

 returned to its original shade on passing into the 

 infra-red. When this effect has once been seen, a 

 very similar play of pale mauve and pink diffrac- 

 tion colours can be discerned on examining the 

 wings of M. achilles itself, and of certain similar 

 insects. 



Very brilliant colours are shown by scale-bear- 

 ing beetles or weevils, like the Brazilian Diamond 

 Beetle (Entimus imperiaJis) (Fig. 3). Michelson 

 admits these colours to be an exception to the 

 general rule, by which he attributes all insect 

 colours to selective metallic reflection, or surface 

 colour. He believes the colours of these beetles 

 to be due to gratings within the scale itself, since 

 as soon as a fluid can enter the scales through 

 a rent or teat the colour vanishes. Moreover, 

 since the light is concentrated in a single spec- 

 trum, he is obliged to assume that the grating 

 has bars, which are asymmetric or prism-shaped, 

 so that they refract the incident rays in the same 

 direction as the diffracted rays of the lateral 



NO. 2658, VOL. 106] 



spectra. For several reasons it is difficult to believe 

 that such saw-tooth-shaped gratings are responsible 

 for the total colour effect. For instance, the very 1 

 saturated complementary colours seen by trans- ' 

 mitted light, and the monochromatic character of 

 the reflected colours at different angles, demand a 

 form of grating structure even more complicated 

 than that described by Michelson, such, for 

 instance, as that named by Prof. R. W. Wood 

 the "echelette grating." Moreover, though as a 

 rule no structures are seen, a very well-defined 

 stratification, 2 (Fig. 2), sometimes appears in cut- 

 ting sections of certain scales, as in the pink 

 weevil, Hypomeces squamosus, var. durulentus. 

 This stratification can be seen in plan, 4 (Fig. 2), 

 and appears to exist throughout the scale, giving 

 it in section the crossed appearance of the strings 

 of a tennis racquet. It seems probable that such 

 a structure would contribute a large share to the 

 total colour effect. Further, a suitable irregu- 

 larity in the periodicity or thickness of the plates 



Fig. 3. — The Rraziliin Diamond Beetle 

 {Kntimus imfierialis), a large, iridescent 

 weevil. The black pits on the wing, 

 cases are lined with gcm-like scales. 

 (Natural size.) 



would account for the existence of the very satu- 

 rated colours of some weevils, and of the very 

 pale and desaturated colours of others. 



Dispersion of Light by Prisms. 



If Michelson's hypothetical prism- or saw-tooth- 

 shaped gratings are omitted, no case of prismatic 

 structure has been met with. It is true that 

 Dr. H. Gadow has explained how the colours of 

 certain feathers might be the result of the roughly 

 prism-shaped structure of the barbules. He 

 supposed that these were placed in such a way, 

 in respect to each other, that each barbule ob- 

 scured part of the spectrum formed by the pre- 

 ceding one, so that partially monochromatic 

 colours would result. Numerous theoretical and 

 practical considerations, however, make this sug- 

 gestion highly improbable. 



The Scattering of Blue Light due to Small 

 Particles. 



The investigations of the late Lord Rayleigh, 

 and others, have shown that the blue of the sea, 

 sky, snow and even tobacco smoke is caused by 

 particles which, being very small compared with 

 the wave-length of light, scatter the blue waves 

 to a much greater extent than the longer red 

 waves. Several colours can be produced in this 

 way, as, for instance, the blue, green, and 

 purple of certain feathers, which are matt, and 

 do not change colour with the angle of incidence. 



