vu : FLIGHT 18g 
the front. The difference is far more apparent when 
we take the whole wing. Above, a birds wing is con- 
vex, so that it passes easily through the air; the 
under-surface is concave and lays hold of it. I am 
referring mainly to the part of it which is nearest to 
the body and which forms a kind of irregular cup, its 
hinder side gently sloping away. Experiments have 
been made with a view to measuring the resistance of 
the air to concave surfaces, but the results do not help 
us much. The irregularity of the wings has not been 
reproduced, the velocity has been uniform, whereas 
that of the wing is very different at different stages of 
the stroke, and no account has been taken of the 
variation of the curves during flight owing to the 
elasticity of the great feathers. 
The duty of the near part of the wing is to a great 
extent that of a parachute. Between the strokes, the 
bird drops, and were it not for these umbrella-like 
supports, the drop would be greater than it is. The 
work of propelling is, as we have seen, done mainly 
by the extremities of the wings, which move with far 
greater rapidity. It is possible to find approximately 
the centre of the action of the air—to find a point in 
the wing so situated that the air shall act with equal 
force on the near and far sides of it. If the wing were 
a triangle, this point would be in a line drawn from 
the middle of the base to the apex, 2 of its length 
from the shoulder. If it were a rectangle, it would be 
at a point on the corresponding line } of its length 
out. Professor Marey estimates that in the actual 
wing it is at a point about 3 of its length from the 
base. “It must vary much in wings of different shape. 
