BIRD LIFE 233 



when the bird painted with it splashes itself; a yellow 

 and a green, the last containing no copper but a good 

 deal of iron. All colour effects seem to be produced by 

 combinations of these, mixed or laid one over another, 

 with or without the help of surface cuttings and 

 polishings. Strange as it may sound when we think of 

 the numbers of Kingfishers and other blue birds to be 

 found almost all over the world, ' blue has not yet been 

 discovered as a pigment.' 



The shot metallic colourings of our English 

 Starlings and of the more gorgeous tropical birds are 

 due to surface chisellings on the feathers, which 

 (Nature is not wasteful of labour) are to be found only 

 on the parts of the feathers exposed to view. 



Without quoting at length Professor Gadow's article, 

 which is a model of condensation, it is impossible to do 

 justice to his subject. But a few lines extracted from 

 the section dealing with 'Structural, Prismatic, or 

 Metallic Colours ' are enough to give some idea of the 

 wonders of contrivance described. These prismatic 

 ' colours change,' he writes, ' according to the position of 

 the observer,' and ' they always change in the order of 

 those of the rainbow.' They are restricted, as a rule, to 

 one particular part of the web, ' the metallic portions 

 of which are composed of one row of compartments, 

 which often partly overlap each other like curved tiles. 

 In the inside black or blackish-brown pigment is 

 collected; and each compartment is covered with a 

 transparent colourless layer of extreme thinness, e.g. 

 0-0008 mm. in Sturnus' (the Starling family). 'The 



