302 



NATURE 



[November 3, 192 1 



Letters to the Editor. 



[T/ie Editor does not hold himself responsible for opinions 

 expressed by his correspondents- Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



Metallic Coloration of Chrysalids. 



' The chrysalids of many butterflies have the appear- 

 ance of being gilded. The area covered by the gilding 

 is variable even in the produce of the same batch of 

 eggs, in some cases showing only round the promin- 

 ences on the back, and in others diffused over the 

 greater part of the surface. 



Modern naturalists seem inclined to explain animal 

 coloration as being either protective, warning, mimetic, 

 or the result of sexual selection, but the ornamental 

 gilding above referred to can scarcely fall into any of 

 these classes. Protective resemblance, i.e. the simi- 

 larity between an animal and its usual surroundings, 

 is met with everywhere, and natural selection would, 

 or might, act continuously, to promote this resemblance. 

 The opportunities, however, for selection to cause 

 members of different genera to assume the same 

 appearance would be much fewer. 



I do not know of any evidence which would point 

 to the models as being of older types than the sup- 

 posed copyists, but even if such evidence existed, the 

 probability that the similarities are due to selection is 

 of a different, and lower, order than that which con- 

 nects selection and congruity of colouring with en- 

 vironment. 



Considering that, apart from the latter tendency, 

 nothing whatever is known as to the origin of the 

 various types of markings and colour patterns, it 

 would seem justifiable to put down many cases, at 

 least, of what is called mimicry tc the action of the 

 same unknown cause. 



Chrysalids are often protectively coloured (amongst 

 the Pieridae, for instance, if the pupa is placed on a 

 wall or paling it is greyish with dark specks, but if 

 on a leaf the colour is a uniform green) ; there are 

 very few positions, however, in which a bright 

 metallic object would not be more conspicuous than 

 a duller surface similar in tint to its background. 



The nature of the metallic colours, of which there 

 are such abundant instances in the animal world, has 

 aKvays been a matter of interest, and during the last 

 twenty years I have examined many hundreds of 

 examples of the kind, and have found in every case 

 that the metallic appearance was due to some form 

 of "interference." 



Co'our, in general, is the result of a selective action 

 of matter on the light which reaches it, and may 

 arise either from a relation between the wave period 

 of the light and the molecular periods of the matter, 

 or from a relation between the wave-lengths of the 

 light and some structural dimension in the substance 

 built up of numbers of molecules. 



The first of these alternatives includes all pig- 

 mentary colours, while the action of the second is 

 properly described as "interference.'' 



Pigmentary colours are analogous to the pheno- 

 mena of resonance in sound, and interference effects 

 to those of combined echoes. 



The interference colours of birds and insects take 

 so many forms that it is not always easy, at first 

 sight, to distinguish them from pigments, but a 

 decisive test may be found by subjecting the coloured 

 material to pressure. 



If the application of pressure causes the colour to 

 disapf>ear, interference may be safely inferred. If 



NO. 2714, VOL. 108] 



the colour changes, but does not vanish, interference 

 is probable, but not certain. 



Fig. I shows the apparatus which I made for 

 carrying out this test. It consists of a quartz plate A 

 and a plano-convex quartz lens B, which can be 

 pressed into contact or withdrawn from A by the 

 adjusting screw C. Other adjustments are provided 

 for traversing A in its own plane and for clamping 

 the apparatus in its correct position on the stage of 

 the microscope. 



In use, B is first brought into contact with A, and 

 the series of Newton's rings thus formed is centred 

 in the field of view. B is then withdrawn and a small 

 piece of the material to be tested is placed on A, and 

 A is moved by its own adjustment screws so as to 

 bring the test-piece into the same position as was 

 occupied by the centre of the ring system. 



Various other tests, such as immersion in various 

 fluids, may be applied to determine the character of 

 the interference, and some of these are mentioned in 

 the paper on "Iridescent Colours" (Proc. Roy. Soc, 

 A, vol. 85, pp. 598 et seq.). 



In the majority of the cases which I have examined 

 the colours seem to be analogous to those of Lipmann 



Fig. I. — Section and plan of apparatus for appl>ii)g pressure to 

 coloured test-pieces. 



films, but it is almost impossible to cut sections thin 

 enough to show this by direct evidence. It may be 

 stated, however, that in the case of birds and insects 

 they are not due to diffraction (an explanation which 

 has often been put forward in books on natural his- 

 tory). I have not yet examined the iridescent colours 

 of fishes and reptiles. From their appearance I should 

 expect that of the former to be interference pheno- 

 mena, but that the splendid iridescence shown by 

 the scales of some pythons may be the result of 

 diffraction. 



It is true that the scales of many Lepidoptera are 

 traversed by fine and regularly spaced lines, and I 

 have used a single scale from a Morpho as a diffrac- 

 tion grating, but the diffraction spectra have fio rela- 

 tion to the colour which is directly reflected. 



To return to the gold of the chrysalids, I find 

 (i) that the gilding disappears completely when pres- 

 sure is applied; (2) no immediate change occurs on 

 immersion in water, alcohol, or xylol, but in the 

 course of a few days the gold tends to become more 

 orange ; (3) the metallic appearance vanishes when 



