DETERMINATION OF NEURAL CONNECTIONS
183
could be modulated to the specificity of its peripheral connection, and
the perikaryon could reselect its central synapses in accordance with its
newly acquired specificity. The concept of myotypic modulation of
Weiss (1941, 1950) is the keystone of the whole hypothesis; in addition
it has the merit of being testable under rigidly controlled experimental
conditions in a model of sufficient simplicity—i.e., the monosynaptic
reflexes of muscle receptors. Unfortunately, the outcome of these types
of tests performed by Eccles and his associates (Eccles, 1964) was not
entirely satisfactory, although there was some evidence of specific changes
of synaptic connections in motoneurons of young kittens whose nerves
had been cross-united during the first days of life.
In spite of its general appeal, the theory meets certain difficulties not
obvious at first sight but becoming more puzzling after further consideration. Although some electrophysiological (Gaze, 1958; Lettvin et al,
1959) and histological (Attardi and Sperry, 1963) support recently has
been found, the specificity concept is based on behavioral experiments.
It has been necessary to assume the existence of various reflex arcs in
the investigation of animal behavior, and from the changes inflicted upon
the reflexes by experimental surgery it has been indirectly deduced that
the appropriate connections of nerve cells in any particular reflex arc are
due to a qualitative difference, presumably of biochemical or physicochemical character, between the neurons. Behavioral observation, obviously, cannot satisfactorily explore the nervous mechanism underlying
different reflex responses, and, apart from that, there are many reasons
pointing toward the operation of different and more general mechanisms.
There are several data in the literature which indicate differences in
the biochemical or physicochemical makeup of neurons. Examples are
the selective action on the nervous system of various viruses, bacteria,
hormones, drugs, and other chemicals, as well as the different metabolic
rate or glutamic acid concentration in neurons belonging to anatomically
distinct regions. The biochemistry of a neurosecretory cell is obviously
different from that of a chemoreceptor cell, whereas the latter differs
from a pyramidal cell, and so forth. These differences may, certainly,
have functional significance. The scale of differences in various qualities
of neurons is fairly large, but cells with common or unlike qualities do
not necessarily participate in the same or different nervous activities.
No specific differences have been shown thus far that could characterize
the neurons of any given pathway or reflex arc; even less is known of
differences that could separate neurons of the same kind from one another
according to their specific locations in the whole group. However, this
does not exclude the possibility of the existence of such functionally
specific qualities.
183
could be modulated to the specificity of its peripheral connection, and
the perikaryon could reselect its central synapses in accordance with its
newly acquired specificity. The concept of myotypic modulation of
Weiss (1941, 1950) is the keystone of the whole hypothesis; in addition
it has the merit of being testable under rigidly controlled experimental
conditions in a model of sufficient simplicity—i.e., the monosynaptic
reflexes of muscle receptors. Unfortunately, the outcome of these types
of tests performed by Eccles and his associates (Eccles, 1964) was not
entirely satisfactory, although there was some evidence of specific changes
of synaptic connections in motoneurons of young kittens whose nerves
had been cross-united during the first days of life.
In spite of its general appeal, the theory meets certain difficulties not
obvious at first sight but becoming more puzzling after further consideration. Although some electrophysiological (Gaze, 1958; Lettvin et al,
1959) and histological (Attardi and Sperry, 1963) support recently has
been found, the specificity concept is based on behavioral experiments.
It has been necessary to assume the existence of various reflex arcs in
the investigation of animal behavior, and from the changes inflicted upon
the reflexes by experimental surgery it has been indirectly deduced that
the appropriate connections of nerve cells in any particular reflex arc are
due to a qualitative difference, presumably of biochemical or physicochemical character, between the neurons. Behavioral observation, obviously, cannot satisfactorily explore the nervous mechanism underlying
different reflex responses, and, apart from that, there are many reasons
pointing toward the operation of different and more general mechanisms.
There are several data in the literature which indicate differences in
the biochemical or physicochemical makeup of neurons. Examples are
the selective action on the nervous system of various viruses, bacteria,
hormones, drugs, and other chemicals, as well as the different metabolic
rate or glutamic acid concentration in neurons belonging to anatomically
distinct regions. The biochemistry of a neurosecretory cell is obviously
different from that of a chemoreceptor cell, whereas the latter differs
from a pyramidal cell, and so forth. These differences may, certainly,
have functional significance. The scale of differences in various qualities
of neurons is fairly large, but cells with common or unlike qualities do
not necessarily participate in the same or different nervous activities.
No specific differences have been shown thus far that could characterize
the neurons of any given pathway or reflex arc; even less is known of
differences that could separate neurons of the same kind from one another
according to their specific locations in the whole group. However, this
does not exclude the possibility of the existence of such functionally
specific qualities.