All the feathers seem equally liable to the defects, though they are more 
conspicious on the commercial wing and tail feathers than on the small 
body-feathers. The most remarkable feature, however, and the one most 
puzzling to account for, is the extent to which variation is shown among 
the feathers growing under the same conditions, not only of different birds, 
but even on the same bird. Single wing plumes may be altogether free 
from the detects, while others, and especially the feathers towards the free 
end of the wing, will be freely barred. Sometimes the feathers of one 
wing will be affected, while those of the others are free. It is also found 
that single feathers growing out of time, that is, not as part of a full 
crop, are much more likely to be deeply barred than those constituting 
part of a full clipping. This apples more especially to single feathers 
FIGS. 5 AND 6. 
Two partly growns-plumes showing very conspicuous bars. In the unexpanded portion 
are shown the deep notches or wrinklings on the feather-sheath which by constricting the 
growing feather produce the bars. Such deep bars occur at intervals of four or five day 
growth, not at each night ring ; smaller notches, not seen in the photograph, represent the 
daily rings. 
which are started in advance of the full crop, than to odd feathers 
appearing after a crop has begun its growth. Lastly it is well recognised 
that the plumes of some strains of birds have a great tendency to 
faultiness, while those of others are nearly perfect; and also that at one 
time the entire clipping from a bird will be barred, while at another the 
crop will be free from them. It is these many aifferences in the occur- 
rence of the bars which render a simple solution of the problem peculiarly 
difficult. 
THE IMMEDIATE CAUSE OF THE Bars. 
The immediate cause of the bar is not far to seek. If one examines 
the lower unopened part of a growing feather in which the faults occur, 
it will be seen that the outer sheath is more or less deeply wrinkled or 
notched at intervals, and by stripping away the sheath it is found that 
