vil FLIGHT 205 
so rotates in the course of the stroke that further 
bewilderment is apt to arise. When the wing is 
folded, what is really the upper side of the humerus 
looks partly upward, but mainly towards the body. 
It may be easily recognised by an unmistakable 
landmark, the foramen or aperture through which the 
bone is aerated (see Fig. 56). In the primitive an- 
cestors of birds, with their forelimbs not yet adjusted 
for flight, this would have been on the upper surface, 
and postaxial. On what is really the under surface, 
but in existing birds looks forward instead of down- 
ward when the wing is at rest, there is another land- 
mark. Where the broad expansion at the rear end 
of the boné begins to narrow down, there may be seen 
a rather long mark on the bone (GP). This is 
where the Great Pectoral muscle attaches—on what is 
-properly the under side near its praaxial edge. In 
describing what we may call the geography of the 
bone, we must always state what would be the case 
if it were set as it is in other animals, and as it is in 
the bird itself during the down stroke. 
At its near end it is broad, and, according as a 
muscle attaches at one margin or the other, can be 
made to rotate either way. A pull from below on the 
preaxial margin will make it turn its under surface 
backward. A pull on the preaxial margin of the 
upper side will make it face forwards. And so forth. 
The Great Pectoral muscle springs from the lower 
part of the keel (supposing the bird is placed breast 
downwards), partly also from the sides of the breast- 
bone above the keel and from the ribs where they 
join the breastbone. From the breast it passes forward, 
