1] 
of the feather-sheath. As the wrinklings are rarely produced uniformly 
they afford a satisfactory explanation of the variations in position and 
extent of development of the faults as compared with the regularity which 
would be expected from the day and night differences alone. 
It results from the above that two conditions are necessary for the pro- 
duction of shrinkage bars in ostrich feathers. First, the alternation of day 
and night growth, due toa slight difference of blood-pressure as the feather 
ws forming, giving rings of different thickness or density; second, greater 
variations in the blood-pressure within the medulla of the feather, leading 
to a partial collapse or wrinkling of the feather-sheath at the weaker night 
rings, whereby the feather is prevented from attaining its full development 
at those particular places. 
The ordinary alternation of day and night growth alone sometimes gives 
a feeble indication of barring, but I doubt if it would ever be so serious 
as to call for the consideration of the practical farmer; when, however, 
it is supplemented by the wrinkling at the places of less density we have 
a combination of conditions capable of producing all the observed 
defects. The indentations stop the proper nutrition of the feather at 
that particular level, and thus greatly intensify the weakening of the 
growth from the lowered night pressure. The combination is one of 
reduced nutrition acting as a primary agent supplemented by a secondary 
mechanical effect, the wrinkling. Both may, and probably do, take place 
at one and the same time; but we may have the night ring without any 
bar. The factors which produce a night ring will give rise to a bar when 
continued far enough. 
Evidently the variations in the blood-pressure and the responses to it 
vary greatly in individual feathers, even among those growing over the 
same part of the body, and it is this which accounts for the haphazard 
character of the appearance of the bars alluded to earher. The variation 
is difficult to understand, but can be partly explained by the different 
stages of feather-growth represented at one and the same time in the 
plumage as a whole, and also by the differences in exposure of the 
individual feathers. Odd feathers are usually more exposed than those: 
forming part of a full crop, and are always more likely to be barred. 
Sometimes where a number of feathers are all at the same stage of growth, 
as in the case of wing and tail feathers after artificial plucking, it is 
found that a bar will be formed at the same level in all the growing 
plumes, showing some marked physiological change in the bird as a 
whole. But usually the barring in individual feathers is quite 
independent, a fact which adds greatly to the difficulty of overcoming it. 
The place of occurrence of the bars and their extent across the flue 
have also been stated to vary much in single plumes. In the shrinkage, 
consequent upon the reduction of pressure, it can easily be understood 
that the wrinklings are not always produced at exactly the same place 
round the weaker night rings, nor do they always assume exactly the 
same form and depth; sometimes one deep indentation will occur at a 
ring, while at others there may be several; some will be on one side of 
the feather, and some on the other; some partly encircling, others almost 
completely so. As a result the flue on opening out is differently affected 
at different levels. 
Wuy tur Bars are Atways Narrow. 
It has been stated that the shrinkage bars are always very narrow 
in their vertical extent, and this now adinits of a simple explanation. 
They are the same vertical height as the night rings. The latter are always 
small compared with the day rings, as the feather grows but little during 
the few hours of the night, and it is only the narrow night part of the 
