yellowish bronze at the tip; others are yellowish green at the 
base to greenish blue at the tip; while some feathers of the purple 
grackle shade from reddish purple at the base to blue, greenish 
blue, green, yellow-green, to yellow-bronze at the tips. These 
sequences of shades are undoubtedly produced by a gradual 
variation in thickness of the color-producing films, for the colors 
adjacent on the feather are always adjacent in Newton’s series, 
and, moreover, the colors shift with increasing incidence, just 
as do Newton’s rings, the shift or change in position of a given 
color being toward the thicker part of the film. 
The grosser structure of the feather may influence the appear¬ 
ance of the iridescent colors very markedly. Not only the 
perfection of the surface formed by the overlapping of the distal 
barbules, but also the visibility of the color, are dependent on 
this. For example, the blue-green iridescent barbules of the 
peacock feather are present in the black center of the “eye;” 
but are arranged so that they are seen edgewise and their color 
ordinarily does not appear. The iridescent colors of many 
birds, particularly humming birds, are often visible only from a 
position in front of the bird, and are quite invisible when the head 
of the bird is pointed away from the observer. This, again, is 
due to the barbules which are so arranged that, instead of having 
their planes in the general plane of the feather, they are skewed 
or tilted somewhat, and lie normal to a direction toward the head 
of the bird, rather than in the plane of the feather. Conse¬ 
quently, they are seen edgewise from other positions, and give 
little or no color. 
One marvels at the way in which these colors are produced 
with such uniformity on different feathers and different birds 
through successive generations, in spite of the fact that a very 
slight variation in the thickness of the films would alter 
the color effect greatly. The production of iridescence seems to 
present a most remarkable case of the regulation of an essential 
structure within very sharply defined limits, not only in numer¬ 
ous individuals but also through many successive generations. 
Perhaps more remarkable still is the case of the white peacock. 
We had hoped that the tail feathers of the white peacock differed 
from those of the ordinary peacock only in not having the dark 
pigment present, in which case dyeing the feather or painting 
the back of it black would bring out the brilliant metallic colors. 
This does not happen because the tail feathers of the white pea¬ 
cock do not have at all the same structure as those of the ordi¬ 
nary peacock. In spite of the absence of color and the complete 
difference in structure, the general pattern of the peacock feather 
has been preserved in the white peacock even though the effect 
is produced in an entirely different way. This must have 
some great biogenetic significance, though fortunately that is not 
a problem for the chemist. 
When this work was begun, the committee believed strongly 
in Michelson’s theory of selective reflection; but it was soon 
found that this theory could not be applied satisfactorily to the 
phenomena observed, and the conclusion was reached finally 
that Michelson was wrong and Lord Rayleigh was right. The 
(4) 
