chap, viii.] METALLIC COLOURS. 273 



well seen in the wings of various insects, the scales of 

 which are marked by extremely delicate strise, or covered 

 by a thin membrane. The rays of white light suffering 

 interference are broken up into their constituent parts, 

 and different colours are produced in different positions. 

 Similar phenomena are to be observed in the shells of 

 Lamellibranchs. The causes of the metallic colours of 

 birds has been carefully investigated by Gadow, who 

 points out that if we look at a feather in a direction 

 nearly parallel to its plane, having one eye between it 

 and the light, it appears black, as it does also when 

 placed between the eye and the light. If we keep the 

 feather steady, and move the eye from one to the 

 other of the just mentioned positions, we notice the 

 gradual appearance of all the metallic colours that the 

 feather is able to display. It is important to observe 

 that these colours do not appear at random, but that 

 the first to be seen are those that are nearest the red 

 end of the spectrum, and the last those that are nearest 

 the violet. No metallic feather ever exhibits a brown 

 or grey appearance, or, in other words, any colour 

 that is not spectral. These facts lead to the belief 

 that the changeable metallic colours are due to a 

 structure comparable to that of a prism ; this structure 

 is formed by a transparent sheath of remarkable 

 thinness (0-00085 mm. in Sturnus, 0-0022 mm. 'in 

 Galbula), which is either perfectly smooth and 

 polished, or has fine longitudinal ridges or numerous 

 minute dots on its surface. Below the sheath there 

 is a brown or dark pigment. As a very small part 

 of the orbit of a curve may be treated as a straight 

 line, the sheath may be regarded as consisting of a 

 number of small prisms ; the reason why colour is not 

 seen when the eye is between the object and the light 

 is that such a prism only produces colour on the side 

 farthest from the light, and therefore refracts no light 

 in the direction of the observer, 

 s 16 



