DEVELOPMENT OF INNERVATION IN TETRAPOD LIMBS 109
It thus appears that in the mammal, and probably in other amniotes,
the partitioning of the lateral motor column into ventral horns, together
with the main period of cell degeneration therein, precedes the further
subdivision of each horn, and the development of motility within the
limbs. This latter phase corresponds to the peak of degeneration in the
anuran ventral horns, which originate as separate centers and undergo
little or no further subdivision. In both, however, is seen a proximodistal
order of events.
To what extent there is cell turnover among the motor cells of the
developing amniote cord is at present an open question. In the chick
embryo, Hamburger (1958) states that by the fifth day the number of
ventral horn cells is already 90% of the final total. Yet Fujita (1964)
has observed that in an embryo injected with tritiated thymidine at 4
days and fixed 3 days later, the ventral horn contains a high proportion
of labeled cells. Between these two stages, there must have been a
synthesis of labeled deoxyribonucleic acid, the mitotic division of cells
thereby marked as well as their migration laterally. These observations
suggest a rapid turnover of cells at this relatively early stage in the
differentiation of the ventral horn. In the mouse fetus, the counts obtained by Harris (1965) of the numbers of viable and degenerating cells
in the lateral motor column do not suggest any large-scale replacement
during this whole period, though an independent estimate of the duration
of degeneration was not made.
Whether there is any subsequent process of selection in the final stages
of differentiation of nerve-muscle relationships in the mammalian fetus
is again not known. One hint has, however, been given by Gamble
(1966), who has observed that in the ulnar nerve of the human fetus
between the third and fifth months, degenerating axons are frequently
seen among healthy fibers, in such juxtaposition that the possibility of
their disintegration resulting from inadequate fixation is excluded.
The tentative and incomplete nature of this sketch of how the vertebrate limb is innervated will need no emphasis. The largest gap is the
absence of any knowledge concerning the development of the synaptic
relationships of the ventral horn cell itself. Hamburger et al. (1966)
have recently shown that in the chick embryo the spontaneous motility
of the hind limbs up to the seventeenth day is little affected by isolation
of the lumbar ventral horn cells, not only by transection of the spinal
cord in the thorax, but also by complete removal of the alar plate of the
cord, together with the rudiments of the lumbar ganglia. It is unlikely
that the development of limb function in a vertebrate larva could proceed
It thus appears that in the mammal, and probably in other amniotes,
the partitioning of the lateral motor column into ventral horns, together
with the main period of cell degeneration therein, precedes the further
subdivision of each horn, and the development of motility within the
limbs. This latter phase corresponds to the peak of degeneration in the
anuran ventral horns, which originate as separate centers and undergo
little or no further subdivision. In both, however, is seen a proximodistal
order of events.
To what extent there is cell turnover among the motor cells of the
developing amniote cord is at present an open question. In the chick
embryo, Hamburger (1958) states that by the fifth day the number of
ventral horn cells is already 90% of the final total. Yet Fujita (1964)
has observed that in an embryo injected with tritiated thymidine at 4
days and fixed 3 days later, the ventral horn contains a high proportion
of labeled cells. Between these two stages, there must have been a
synthesis of labeled deoxyribonucleic acid, the mitotic division of cells
thereby marked as well as their migration laterally. These observations
suggest a rapid turnover of cells at this relatively early stage in the
differentiation of the ventral horn. In the mouse fetus, the counts obtained by Harris (1965) of the numbers of viable and degenerating cells
in the lateral motor column do not suggest any large-scale replacement
during this whole period, though an independent estimate of the duration
of degeneration was not made.
Whether there is any subsequent process of selection in the final stages
of differentiation of nerve-muscle relationships in the mammalian fetus
is again not known. One hint has, however, been given by Gamble
(1966), who has observed that in the ulnar nerve of the human fetus
between the third and fifth months, degenerating axons are frequently
seen among healthy fibers, in such juxtaposition that the possibility of
their disintegration resulting from inadequate fixation is excluded.
The tentative and incomplete nature of this sketch of how the vertebrate limb is innervated will need no emphasis. The largest gap is the
absence of any knowledge concerning the development of the synaptic
relationships of the ventral horn cell itself. Hamburger et al. (1966)
have recently shown that in the chick embryo the spontaneous motility
of the hind limbs up to the seventeenth day is little affected by isolation
of the lumbar ventral horn cells, not only by transection of the spinal
cord in the thorax, but also by complete removal of the alar plate of the
cord, together with the rudiments of the lumbar ganglia. It is unlikely
that the development of limb function in a vertebrate larva could proceed