August i8, 1923] 



NA TURE 



243 



proposed : (i) Inclined to zero and left, the reading of 

 the level bubble at the end of the first swing is taken as 

 the hardness number. The softer the material, when 

 the indentation due to the weight of the instrument is 

 deep, the shorter is the swing. (2) The time period of 

 an oscillation is another measure of hardness. The 

 time in making ten swings is taken as the hardness 

 number. Thus the time of ten swings on glass is 

 100 sec, on hardened steel 50 to 85 sec, on soft 

 steel 20 to 40 sec, on lead 3 sec. The pendulum 



is set in oscillation through a small arc by the touch 

 of a feather. The sensitiveness of the instrument is 

 very great, and it gives definite indications with the 

 hardest materials. 



Dr. Stanton has designed an ingenious instrument 

 in which the deformation of a very hard ball used in 

 the indentation test is substituted for the deformation of 

 the material. This gives a much opener scale for hard 

 materials. But the instrument is one for laboratory 

 rather than workshop use. W. C. U. 



; Structural Colours in Feathers.^ 



By Prof. Wilder D. Bancroft. 



IN pigment colour we have absorption of light due 

 to the molecular structure of the substance under 

 observation. We speak of structural colours when 

 the observed colour is due to, or is modified strongly 

 by, the physical structure. Typical cases of structural 

 colour are observed with prisms, diffraction gratings, 

 thin films, and turbid media. In the case of feathers 

 we find that the blacks, reds, oranges, yellows, and 

 browns are pigment colours, but that the ordinary 

 blues and greens are not blue and green by trans- 

 mitted light, and that the so-called metallic or irides- 

 cent colours, such as those of the peacock, are structural 

 colours. 



Biologists have often talked of prismatic or diffrac- 

 tion colours, apparently because those were the only 

 structural colours that they knew about ; but they 

 have never tried to show that any arrangement of 

 prisms or gratings would give the actual colours 

 observed. Since prisms and gratings give no colour 

 in a uniform diffused light, it is only necessary to look 

 at a feather on the north side of a house, prefer- 

 ably on a grey day, and all prismatic or grating 

 colours will disappear. Nothing of the sort happens, 

 except to an almost neghgible extent, with some 

 moths. 



If we have a turbid medium with fine particles, the 

 scattered light is predominantly blue — Tyndall blue — 

 and the transmitted light is reddish. Familiar 

 examples of this are skimmed milk and cigarette 

 smoke. The blue of the sky is also a Tyndall blue, 

 the scattering being due in large part, however, to 

 the molecules of nitrogen and oxygen, as was shown 

 l)y the late Lord Rayleigh. In feathers of the non- 

 iridescent type, Haecker showed that we have myriads 

 of tiny bubbles in the horn which scatter the light, 

 and a black backing which cuts off all transmitted 

 light. On filling the bubbles with a liquid having 

 approximately the same index of refraction as the horn, 

 the scattering ceases and the blue colour with it. 

 On putting in carbon bisulphide, which has a much 

 higher index of refraction than the horn, the blue 

 reappears because we again have a turbid medium. 

 The blue of the feathers can be reproduced wonder- 

 fully by heating a hard glass tube until it begins to 

 devitrify. The myriads of small crystals which are 

 formed scatter the light, and a beautiful blue is obtained 



»I Synopsis of a lecture delivered at University College, London, on 

 June I, at the University of Aberdeen on June 7, and before the 

 Manchester Literary and Philosophical Society on July 19. 



NO. 2807, VOL. 112] 



if the inside of the tube is coated with a black varnish 

 to eliminate transmitted light. 



In almost all cases of non-iridescent green feathers, 

 there is no green pigment and the effect is due to the 

 superposing of a yellow pigment on a structural blue. 

 This can be shown in a number of ways. If we take 

 a green feather and boil it long enough in alcohol, 

 the yellow pigment dissolves and the feather turns 

 blue. If we expose the green feather long enough 

 to an intense light, the yellow pigment bleaches and 

 the feather becomes blue. If we scrape the surface 

 of the feather with a sharp knife, we can peel off 

 a layer of yellow horn and the feather again turns 

 blue. 



The metallic or iridescent colours, such as those of 

 the peacock, were considered by Rayleigh to be the 

 interference colours of thin films like those observed 

 with oil films on the streets, while Michelson believed 

 that they were so-called surface colours from solid 

 pigments. Fuchsine gives a yellow-green surface 

 colour quite different from the magenta colour by 

 transmitted light. Our experiments have satisfied 

 us that Rayleigh was right and Michelson wrong. 

 There are no bright-coloured pigments in peacocks' 

 feathers or in any feathers of that type. In the case 

 of the peacocks there are triple films, but this is not 

 so in the neck feathers of the pigeon. ♦ 



Nobody has ever extracted any bright-coloured 

 pigment from any iridescent feather, and we have 

 confirmed this, using a large number of organic solvents. 

 The change of colour with the angle of incidence is 

 what it should be for thin films, while magenta shows 

 practically no change of colour with changing angle 

 of incidence if one does not use polarised light. If 

 one swells the feather by exposing it to phenol vapour, 

 the change of colour is what one would predict from a 

 thickening of the film. If one destroys the dark 

 pigment, the colour disappears almost completely, 

 though it can still be seen at certain angles. It can 

 be brought back by staining the feather with a dark 

 pigment. In the white pigeon, the iridescence of the 

 neck feathers is very difficult to see, but it can be 

 brought out vividly by staining the feather. Unfor- 

 tunately the physical structure of the tail feathers of 

 the white peacock is quite different from that of the 

 ordinary peacock, and consequently staining does not 

 develop brilliant colours. 



The average thickness of the films in the iridescent 

 feathers is about 0*5 \i- or 1/50,000 inch. 



