82
ARTHUR
HUGHES
of behavior with histological studies. This approach was first shown of
value in the pioneer work of Coghill (1929) on the origin of the swimming mechanism in the larva of Ambystoma.
There are great advantages
in using developing Amphibia for enquiries into the genesis of limb
function, with regard to accessibility of observation and to experiment,
though it is already clear that amniote development in this respect
proceeds along somewhat different lines, as will be described in a final
section of this article. However, the amphibian studies here described
are on anuran larvae and not on those of urodeles. The reason for this
choice is related mainly to one untypical feature of the latter group
among tetrapods, namely, the absence of differentiated ventral horns
within the somatic motor column of the cord.
Within the spinal cord of an anuran tadpole there are virtually two
nervous systems. One develops early in larval life and is concerned with
axial swimming. Its sensory neurons are the Rohon-Beard (Burckhardt,
1889) and Fritsch cells (Hughes, 1957), and its motor neurons a series
of "primary motor cells." The second neuronal assembly is made up of
a vastly greater number of cells, within the dorsal root ganglia and the
ventral horns of the cord. This system develops slowly, much as in an
embryo, and only becomes functional toward the approach of metamorphosis.
The early phases of development of the larval nervous system show
examples of the growth of individual fibers to precise end stations. This
direct method of production of a relatively simple nervous system is
correlated with the necessity of rapid attainment of function. In amphibian larvae the first ventral roots, each composed of very few fibers, are
collateral branches from a longitudinal spinal tract. These roots leave
the cord at such angles that they innervate myotomes not opposite their
level of emergence but at some distance away. The ventral roots at the
anterior end of the cord run craniolaterally, and elsewhere caudolaterally.
They thus radiate from an anterior segment of the spinal cord which has
been shown to be concerned with the initiation and control of swimming
movements (Hughes, 1959).
In the relatively leisured development of the innervation and function
of a limb, three phases are distinguishable, namely, first a period preceding motility, second, one of rapid development of limb movement and
of its neurological background, and third a final period during which
functional adjustments are made with increasing efficiency in the action
of the limb. The description and analysis that follows is largely based
ARTHUR
HUGHES
of behavior with histological studies. This approach was first shown of
value in the pioneer work of Coghill (1929) on the origin of the swimming mechanism in the larva of Ambystoma.
There are great advantages
in using developing Amphibia for enquiries into the genesis of limb
function, with regard to accessibility of observation and to experiment,
though it is already clear that amniote development in this respect
proceeds along somewhat different lines, as will be described in a final
section of this article. However, the amphibian studies here described
are on anuran larvae and not on those of urodeles. The reason for this
choice is related mainly to one untypical feature of the latter group
among tetrapods, namely, the absence of differentiated ventral horns
within the somatic motor column of the cord.
Within the spinal cord of an anuran tadpole there are virtually two
nervous systems. One develops early in larval life and is concerned with
axial swimming. Its sensory neurons are the Rohon-Beard (Burckhardt,
1889) and Fritsch cells (Hughes, 1957), and its motor neurons a series
of "primary motor cells." The second neuronal assembly is made up of
a vastly greater number of cells, within the dorsal root ganglia and the
ventral horns of the cord. This system develops slowly, much as in an
embryo, and only becomes functional toward the approach of metamorphosis.
The early phases of development of the larval nervous system show
examples of the growth of individual fibers to precise end stations. This
direct method of production of a relatively simple nervous system is
correlated with the necessity of rapid attainment of function. In amphibian larvae the first ventral roots, each composed of very few fibers, are
collateral branches from a longitudinal spinal tract. These roots leave
the cord at such angles that they innervate myotomes not opposite their
level of emergence but at some distance away. The ventral roots at the
anterior end of the cord run craniolaterally, and elsewhere caudolaterally.
They thus radiate from an anterior segment of the spinal cord which has
been shown to be concerned with the initiation and control of swimming
movements (Hughes, 1959).
In the relatively leisured development of the innervation and function
of a limb, three phases are distinguishable, namely, first a period preceding motility, second, one of rapid development of limb movement and
of its neurological background, and third a final period during which
functional adjustments are made with increasing efficiency in the action
of the limb. The description and analysis that follows is largely based