GLIDING 21 
(taking his weight as four pounds and four-fifths) 
his wings would together measure twenty square 
feet! This would be a monstrous acreage of wing 
to raise and lower. But when we have pronounced 
it monstrous, we have still to answer the question 
why it is that the big flyer requires, in proportion 
to his weight, a comparatively very small support- 
ing surface ? 
Let us imagine the Swallow supplied with wing- 
area at no more liberal rate per pound weight than 
that at which the Stork is supplied. Then, taking 
the Swallow’s weight to be about five-sevenths of 
an ounce, he would have five square inches of 
wing-surface—two and a half on either side—a 
miserably poor allowance. A wing so small would 
be largely made up of margin, and the air would 
escape at the edges. The gnat has over four and 
a half square yards of wing for one pound weight. 
His actual allowance for his almost imponderable 
insignificance is considerable, but if we make pro- 
vision for him at the rate at which the Stork is 
supplied, his wing-surface becomes a mere point, 
“without parts and without magnitude,” to quote 
Euclid’s familiar definition. The air would offer 
no resistance to so near an approximation to the 
theoretical point. Here, no doubt, we are getting 
at the main fact that explains the comparatively 
small size (when weight is allowed for) of the wings 
of the great flyers. Since the wing-surface is much 
larger absolutely the air does not escape so easily 
at the margins ; each square inch is more effective, 
since there is less waste. Later on I shall make 
a further comparison of big birds and small, but 
