vil FLIGHT 207 
lies below it. But the humerus turns as on a pivot 
at the joint, and if it be pulled inward it will be 
raised to an upright position, if only the pull comes 
from a point not below the pivot. A tendon, therefore, 
must be made to pass through a pulley fixed at a 
point high enough, and must attach to the inner side of 
bone some little way out from the joint. By such means 
a mast may be raised, if it be prevented from slipping 
at its base. A pulley on a level with the shoulder- 
joint is available. There are two bones, as explained 
above, which meet to form the socket, the coracoid 
and the scapula or shoulder blade. On their inner 
side the top of the clavicle or merrythought meets 
them, and the three together form a tunnel of bone. 
Through this passes the tendon and fastens on to the 
back of the humerus near its preaxial margin. 
The upstroke is at first in a backward direction, 
mainly owing to the action of the wind due to the bird’s 
own velocity ; such a wind must necessarily drive the 
wing backward and at the same time lift it? The move- 
ment to rearward, which is often difficult to detect in 
photography, is clear in diagrams obtained by Professor 
Marey by various methods. Towards the end of the 
upstroke the wing is moved forward, till it stands above 
the pulley and the pivot, mainly by the action of the 
Elevator muscle. The way to see the working of 
these muscles is to take a dead bird and cut away the 
1 The third pectoral muscle, springing from the outer and 
lower part of the coracoid and from the sternum close to the 
ribs, assists—its chief office being to draw the wing backward. 
It attaches to the upper side of the humerus, just over the air 
foramen. : 
2 See p. 221. 
