416 DR. H. GADOW ON THE COLOUR OF FEATHERS. [May 2, 



the green appearance is the result of a mixture of the yellow-pig- 

 ment colour and of a blue optical structural colour. However, this 

 cannot well be always the case, since most green feathers do not 

 show that peculiar structure v/hich we invariably meet with in blue 

 feathers. 



All the green feathers which I have examined show the following 

 structure : — Generally a transparent smoothly surfaced sheath sur- 

 rounds the rami and the radii, which are both green. Between this 

 sheath and the invariably present yellowish, brownish, or pinkish 

 pigment one sees a system of ridges and fine pits. These ridges are 

 shorter and less regular than those observed in yellow feathers, and 

 the little pits are rather irregularly dispersed over the shaft and 

 plane of the barbs and barbules. The more regular and parallel 

 these furrows are, the more approaches the green colour to a 

 yellowish tinge. As we know of no green feathers without any pig- 

 ment, and always with such an irregularly ridged and furrowed 

 surface-structure, we cannot say that this structure directly produces 

 green, nor that it produces blue. We must accept that they break 

 the yellow light, issued from the yellow pigment, into green. 



Red feathers are frequently surrounded with a thick transparent 

 sheath, for instance those of Rhamphastus ; but they have no peculiar 

 or particular surface-structure, and the large wrinkles which we observe 

 in them seem to be merely the result of a drying-up process of the 

 horny feather-substance. In orange or orange-brown feathers, how- 

 ever, we frequently find a dark red pigment and yellow surface- 

 structure. 



Explanation of subjective or metallic colours.— We speak of 

 metallic colours if the feathers under reflected light appear with a 

 metallic gloss, and if their glossy colour changes into another one 

 according to the position of our eye. If we look in a direction nearly 

 parallel to the plane of the feather it will appear black. This can 

 be done in two ways (fig. 3, p. 420) : first, with our eye between the 

 object and the light, a position which I propose to call A ; secondly, 

 with the object between the light and the eye, position C. By 

 passing the eye from A to C, along the line indicated by the arrow, 

 we notice the gradual appearance of all the various metallic colours 

 which the feather is able to display. We further observe that these 

 colours do not appear at random, but, and this is of the greatest 

 importance, that they begin with the colours nearest to the red side 

 of the spectrum, and end with the violet. The position just inter- 

 mediate between A and C is that in which we look vertically down 

 upon the plane of the feather, with the object turned fully to the 

 light ; no matter, however, whether this position is produced by 

 looking at the feather in the way as indicated by diagram B or D. 

 This position we call B. 



In order to ascertain this fact, I have examined, under these three 

 positions, about eighty birds of all orders, wherever metallic colours 

 were present, and I did not find one single exception to the rule. 

 With the exception of two particular cases, which I shall explain 

 later on, the metallic parts of all these birds look perfectly black in 



