A BIRD’S OUTER WRAP. 33 
It is necessary still to shed the epidermis, but 
birds and mammals get rid of it in particles almost 
imperceptibly. Some penguins, however, are said to 
exhibit the old reptilian habit by shedding theirs in 
great flakes with feathers attached. 
It is quite probable that originally the barrel or 
toothpick part of the feather was flat, and that the 
gradual incurving of the edges finally resulted in 
welding it into a tube. This tube thus inclosed a 
glandular core, which nourishes the feather till it cut 
itself off from further fuod by the closing of the lower 
end of the barrel. We sometimes call the dried-up 
cells of this gland “the pith.” 
The tubular form was not necessary till flight be- 
gan, as no strain could previously be put upon the 
quills. But strain demands strength, and the tubular 
form is the strongest arrangement of a definite 
amount of material consistent also with lightness. 
Likewise, until flight came, the development of the 
shaft (the rest of the stem) was not demanded. The 
shaft, however, may have been simply the result of 
the natural growth or general progress of the feather 
beyond the surface. The downs, which appear primi- 
tive now, have practically no shafts, but this may be 
the result of degeneration. 
As flight has influenced so much the structure of 
feathers, the vanes were doubtless at first flossy, and 
perhaps without barbules, or little side vanes upon 
the vanes. These were subsequently added with 
hooks to hold the barbs together into one solid, air- 
resisting surface. The Archwopteryx, the most primi- 
