FEATHERS AND PATTERNS
15
and, as the cylinder lengthens, cut into tissue progressively more dorsal
(Fig. 3). It has been suggested ('Espinasse, 1939) that this results from a
compression of the cylinder in the feather sheath; the possibility of such
compression aligning the barb ridges parallel to the long axis of the
slender down cylinder, while allowing those of the fatter contour cylinder
to deviate from the generators, is still attractive. However, it fails to
account for the always external (dorsal) position of the rachis, and for
those feathers (e.g., wing primaries) whose rachis divides the vane
unequally in a predictable manner. At least no gross deformation of the
circular cross section of the cylinder can be observed in these feathers
(Fig. 6).
In culture the orientation of barb ridges seemed (Cohen, 1957, 1959;
Cohen and 'Espinasse, 1961) to be unrelated to the original orientation of
the expiant in its collar (except for midventral expiants, see below) and
to be determined by accidental factors incident to the handling of the
tissues. Recent studies of the same kind, however (see below), throw some
doubt on this belief. There seems to be no doubt that the change from
generator-parallel to generator-oblique barb ridges is associated closely
with the acquisition by the dermal papilla of a higher mucopolysaccharide level on the dorsal side (Lillie, 1942; and Fig. 6B). Before such
acquisition the dermal papilla is said to stain evenly across its diameter with Alcian Blue. Mucopolysaccharide extract of dermal papillae
has been investigated in Cruickshank's laboratory for its effect on H
3 -
thymidine uptake by guinea pig ear epidermal cells (J. Cohen, C. N. D.
Cruickshank, E. Hell, and P. Friar, unpublished investigations, 1963),
and incorporation rose to 150% of normal with this additive to the usual
nutritive medium (G.C.M.).
2 This seems dramatic, but a similar preparation from embryo extract raised incorporation to nearly 200% of
normal. Furthermore, the constant production by midventral contour
collar of "ventral triangles" (Fig. 7C) (which seemed at first to support
Lillie's view of barb formation) in tissue culture, i.e., remote from the
dermal papilla, argues against such a direct effect of dorsal papilla as
the staining result might suggest. Even if the "ventral triangles" were
produced at the lowest mucopolysaccharide level, it would be expected
that in culture all segments of the collar would produce barbs in this
way, and this is not the case.
There is a further complication. Many feathers possess an "afterfeather": a new rachis appears at the ventral side, and barb ridges align
to it as to the major (dorsal) rachis (see the breast feathers in Fig. 2A).
Its "field" is usually small, however (less than 20% of the circumference
is common, but in the emu the two rachises have equal or nearly equal
2 G.C.M. = 40% horse serum and 60% Hank's solution with antibiotics.
15
and, as the cylinder lengthens, cut into tissue progressively more dorsal
(Fig. 3). It has been suggested ('Espinasse, 1939) that this results from a
compression of the cylinder in the feather sheath; the possibility of such
compression aligning the barb ridges parallel to the long axis of the
slender down cylinder, while allowing those of the fatter contour cylinder
to deviate from the generators, is still attractive. However, it fails to
account for the always external (dorsal) position of the rachis, and for
those feathers (e.g., wing primaries) whose rachis divides the vane
unequally in a predictable manner. At least no gross deformation of the
circular cross section of the cylinder can be observed in these feathers
(Fig. 6).
In culture the orientation of barb ridges seemed (Cohen, 1957, 1959;
Cohen and 'Espinasse, 1961) to be unrelated to the original orientation of
the expiant in its collar (except for midventral expiants, see below) and
to be determined by accidental factors incident to the handling of the
tissues. Recent studies of the same kind, however (see below), throw some
doubt on this belief. There seems to be no doubt that the change from
generator-parallel to generator-oblique barb ridges is associated closely
with the acquisition by the dermal papilla of a higher mucopolysaccharide level on the dorsal side (Lillie, 1942; and Fig. 6B). Before such
acquisition the dermal papilla is said to stain evenly across its diameter with Alcian Blue. Mucopolysaccharide extract of dermal papillae
has been investigated in Cruickshank's laboratory for its effect on H
3 -
thymidine uptake by guinea pig ear epidermal cells (J. Cohen, C. N. D.
Cruickshank, E. Hell, and P. Friar, unpublished investigations, 1963),
and incorporation rose to 150% of normal with this additive to the usual
nutritive medium (G.C.M.).
2 This seems dramatic, but a similar preparation from embryo extract raised incorporation to nearly 200% of
normal. Furthermore, the constant production by midventral contour
collar of "ventral triangles" (Fig. 7C) (which seemed at first to support
Lillie's view of barb formation) in tissue culture, i.e., remote from the
dermal papilla, argues against such a direct effect of dorsal papilla as
the staining result might suggest. Even if the "ventral triangles" were
produced at the lowest mucopolysaccharide level, it would be expected
that in culture all segments of the collar would produce barbs in this
way, and this is not the case.
There is a further complication. Many feathers possess an "afterfeather": a new rachis appears at the ventral side, and barb ridges align
to it as to the major (dorsal) rachis (see the breast feathers in Fig. 2A).
Its "field" is usually small, however (less than 20% of the circumference
is common, but in the emu the two rachises have equal or nearly equal
2 G.C.M. = 40% horse serum and 60% Hank's solution with antibiotics.