HABITS AND COLORATION OF THE STARLING. 289 
of zones of browns, dull greens, reddish whites, &c. The exact 
- relation between each colour and the thickness of the film has 
been ascertained in glass, water, and other substances, but it is 
not known in the case of the ceratin of feathers. 
The difficulties attending the microscopical examination of 
feathers are very great. Inspite of the most careful methods of 
imbedding and cutting, the sections of metallic feathers exhibit 
little of value. It is impossible to learn the nature or the 
extent of the transparent outer sheath, but stained and oblique 
sections show that it is actually present. I have only discovered 
one way of demonstrating the true nature of metallic feathers, 
and that is by the use of solutions of caustic potash. This is a 
solvent of ceratin, and by it we can remove a portion of the thin 
outer film, and as each particular colour depends upon a certain 
thickness of the layer, the general tint of the feather is affected 
at once by the potash, and we get some illuminating results 
from our experiments. 
The comparatively dull colours of the Starling prevent it 
being a convenient subject, and for experimental purposes I 
have found other birds more suitable. Peacock feathers are ad- 
mirable, as are those of Ducks, exotic Pigeons, or Humming 
Birds; and by the exercise of ordinary caution I have turned 
green feathers to brown and red, red to green, and so on. In 
most instances these changed colours are permanent when the 
feathers are dried, and a Humming Bird with an area of golden 
green in the middle of a specific ruby patch, or a tail-feather 
beautifully green where it should be violet, is well worth the 
trouble taken in its metamorphosis. 
In the Starling the potash removes the lustre of green 
feathers, and changes purple feathers into green. Those who 
have not made the acquaintance of the wide range of thin plate 
colours need to be warned that this does not quite prove that the 
critical layer of ceratin is thicker in purple feathers than it is 
in green, but for our present purposes we can assume that it is 
so. Thus it would be possible, in theory, to’change a specimen 
of Sturnus menzbieri into typical S. vulgaris, but in practice this 
would be impossible, unless we treated the plumage feather by 
feather. Here we have an illustration of the dangers that 
attend the student who attaches too much importance to colour 
