igo6. Anderson. — Ani?nal Coloj-ation. 115 
much be5^ond dusk. The same holds for shades of blue and 
several other colours. Thus, pink paper appeared pale pink 
with a light (white) which is noted as "dusk," "dirty" white 
was the colour with still less light, whilst the colour is marked 
as " pale," when all trace of red had disappeared. Smooth 
orange-coloured paper seemed bluish in a " dusk," and soon 
became black. This colour it assumed when pink became 
a dirty white. Rough orange paper appeared even darker, 
whilst green paper was in " dusk " yellow, then in increasing 
darkness white. Red became darkish and even blackish. 
Yellow, which is almost as conspicuous as white, became 
whitish, which was succeeded by pale white. The disadvantage 
of colouring walls red has been long recognized. Red walls 
do not help one to find a gate on a moderately dark night un- 
less the gate be painted w^hite. White, in any case, is much 
better to mark walls or wires, or, failing this, a pale colour. 
When we take a greater series of colours — say, forty fairly 
well known colours — the general facts are the same. Colours 
that in bright light seem so important become less conspicuous 
in the dusk and lose their character before we are unable to 
distinguish between black and white sufficiently well to 
identify printed words that indicate the name. 
Cardinal gets crimson and black, but scarlet gets light 
crimson and then dark red. The darker shades of terra 
cotta become black soon, but the lighter get brown before they 
get black. Navy blue gets soon black. lyight blues become 
dark and gradually black, or pale according to the degree of 
lightness. Thus, palest blue gets first pale blue ; turquoise 
gets pale blue, then pale ; dark green grows black in the fail- 
ing light ; and palest green becomes bluish white and then 
pale. 
When we try to apply observations to animals without 
actually noting the demeanour of the animals, then indeed is 
it difficult to be sure that the results are correct, for the mode 
of aggregation of the coloured particles in the skin, hair, or 
feathers affects the result. It has been proved that the 
superficial layer alone may not produce the full effect 
observed. One notices that the same surface may appear to 
reflect and absorb in an irregular manner, and this may be 
due to the aggregation of the particles. 
A 2 
