86 
«nl)staiico and is vefractcd to ]>, whove it is reflected to K, and then 
Anally I'efracted in the dii'ection 1*1 1). Tt is a well-known fact in¬ 
physics that under these conditions any ray striking the Arst surface 
parallel to A C will be reflected in a direction exactly parallel to E D, 
and that no matter whether the ray comes from the Arst or second 
surface, the ray on leaving the substance will always travel in a direc¬ 
tion parallel to E D, provided it reaches the surface in a direction 
parallel to A C. Arguing from this, let us suppose that F E strikes the 
surface at E, it will therefore be reflected in the direction E D, i.c., 
along exactly the same path as the refracted and I’eflected ray A C B E. 
Now it is clear that the ray A C B E hasfarther to travel than the ray 
F E, and should it arrive at E, half the length of the red light waves 
behind F E, it is clear that when they are both travelling in the direction 
E D and one-half a red wave length behind the other, the whole of the 
red waves will be annihilated, and the result is that we shall see the 
complementary colour, say, the green-blue ; l)ut if, on the other hand, 
the green and blue waves were those annihilated, we should see the red 
or reddish-orange. 
One other point of great importance in connection Avith this matter 
is that, according to the well-known law called the “hiAvof sines,” the 
angle of refraction always bears a Axed ratio to the angle of incidence 
when the same substance is being considered, and this being so, it 
folloAvs that by altering the angle of incidence you alter the angle of 
refraction proportionally, and the net resAilt is that you alter the 
distance A (! B E has to travel in respect to the distance traA'clled in 
the same time by F E. If this is not clear to everyone in the room, 
this diagram, which is drawn to scale, will, I think, make it so— 
Noav it is clear from this diagram that the distance travelled by 
C B E is Ave or six times as great as H E. In the loAver part it is 
clear that F H = A C, and that, therefore, the difference lies betAvoen 
C B E and H E, and as by optical laAvs If E = C K, if C B E is Aa’o 
or six times as long as Cl K it Avill bear the same ratio to II E. Noav, 
in the upper part of the Agure, it is less than twice the distance owing to 
the altered angle of incidence. This alteration shoAvs at once that 
under these different conditions you cut off a different set of AvaA’o- 
lengths, and the result is that the insect appears to change colour. 
This has an important bearing on colour protection, Avhich I will notice 
later on. 
White colours are due to di.spersion of light, /.c., to the numerous 
surfaces continually reflecting liglit in all directions and reflecting all 
light. I cannot do better than quote Boulton’s admirable examples of 
Avhito duo to irregular dispersion, rh., glass Avhe/i crushed to a poAvder 
becomes Avhite from the innumerable surfaces of air and glass Avhich 
