1882.] DR. H. GADOW ON THE COLOUR OF FEATHERS. 415 



the light falling upon and passing through a blue feather is likely to 

 undergo. First, part of the rays will be simply reflected from the outer 

 surface of S S (fig. 1 , p. 4 1 3) ; secondly, the rest, before passing through 

 this stratum, will be variously broken and reflected before reaching 

 the coating ps, since the stratum SS is not homogeneous, but consists 

 apparently of several irregular scales and secondary strata ; thirdly, 

 the coating, ps, breaks the rays again and partly reflects them, and, if 

 it is only O'OOOG mm. thick, as in l^itta, it is thin enough to allow 

 the appHcation of the theory of thin-plate colours ; fourthly, the 

 system of ridges ; fifthly, some rays will reach the layer of brownish 

 pigment. How much of them is absorbed, how much reflected as 

 brownish light, and what the changes are of this brown light before 

 it comes up again to the surface, we cannot tell. Again, the ray a 

 will be under diff"erent conditions to the ray c. To follow and 

 to calculate all these changes would be almost a superhuman task. 

 We know only the result, namely blue colour. 



By the application of the theory of colours of narrow ridges we 

 are enabled to explain several other colours, fortunately under less 

 difficult circumstances. We have seen before that many yellow 

 feathers owe their colour to a yellow pigment. But several of them 

 do not contain any pigment. The thin rami and radii of the downy 

 part of a feather of Pitta, for instance, appear coloured (yellow) only 

 under direct light, but they are colourless if examined under the 

 microscope with transmitted light. Now in yellow feathers, no 

 matter if they contain pigment or not, the surface shows very fine 

 longitudinal ridges, which are more or less parallel to one another, 

 and which appear as straight lines. This I found was the case in 

 the yellow feathers of Pitta, Psittacula, Arachnothera, Picus, and 

 Parus. The distance between the top of two neighbouring ridges 

 varied from O'OOl to 0*0005 of a millimetre. That there are real 

 ridges on the surface we can see on a transverse section of a yellow 

 radius. The radius of a yellow pectoral tuft-feather oi Arachnothera 

 (Plate XXVIII. fig. 6) had a diameter of 0"007 mm. ; as there were 

 about twelve such ridges, like a, ft, y, their distance could not be 



greater than . tt =0"0018 mm. In Pitta the radius of a 



half-downy feather had a diameter of 0'012 mm. All round there 



were about twelve ridges, and the breadth of one ridge was rather 



smaller than the interstices ; therefore the breadth of one ridge 



12-0 

 must be smaller than . 7r = 0'0015 mm. Another method of 



calculating gave 0*00 12 as the breadth of one ridge. 



Violet feathers. — Similar ridges exist on the surface-coating of 

 violet metallic feathers, as, for instance, in Mthopyga and in Sturnus ; 

 but the ridges do not appear to be quite straight, moreover they are 

 much finer ; in Sturnus only 0*00085 mm. 



Green feathers. — Only in the Musophagidse green pigment has 

 been found. All other green feathers contain only either zoofulvin 

 or a black-brown pigment. Krukenberg suggested therefore that 



