60 THE FLIGHT OF BIRDS 
at a moment when they were not held symmetrically. 
Another plan is to V@hd sideways at the waist, 
so as to move the centre of gravity towards one 
side. I was long doubtful whether this plan was 
actually adopted. At length I have obtained a 
photograph of a Gull, taken from below, where the 
waist, thus bent, is clearly shown, but unfortunately 
this photograph will not stand reproduction. Below 
I give some measurements which show that birds 
that are good steerers have greater suppleness of 
waist than clumsy steerers. 
The tail also is undoubtedly used for steering, 
though, I believe, it is more frequently of service 
in maintaining or restoring equilibrium. The rule 
is, as I have shown, that if the legs are long the 
tail is small, and as it is indisputable that long 
legs are very useful for balancing and of but little 
use in steering, we may infer that the tail is more 
a balancer than a rudder. Nevertheless, it is of 
much use in steering. Though its edges look to 
right and left, it can be made an effective rudder 
by the lowering of one side more than the other. 
Web-footed birds probably use their feet occa- 
sionally for steering purposes, but I doubt if their 
action counts for very much. When a Duck, for 
example, is hurtling through the air, if he lowers 
one foot, the resistance of the air will double it up, 
not expand it. It is very different when a Duck 
in swimming kicks backward; the water acts 
upon the under-surface and spreads the webs to 
their fullest extent. Some birds when alighting 
use their feet in the same way. A Gannet, for in- 
stance, kicks hard in order to correct his balance 
