DETERMINATION OF NEURAL CONNECTIONS
213
determined by intrinsic factors. Different sections of the center retain their
morphological character after being transplanted in heterotopic places
(Detwiler, 1943; Straznicky, 1963), or following complete isolation from
the rest of the nervous system (Hamburger, 1946; Bueker, 1943; Reddick,
1945; B. S. Wenger, 1951). This demonstrates a self-differentiation capacity of the parts. The early belief that the adult amphibian nervous
system is capable of reconstituting extirpated regions of the brain or cord
has not gained support (Lorente de No, 1921; Piatt, 1955). The remarkable capacity of the retinal pigment epithelium to regenerate regionally
specific neural retina (Stone, 1950) may be regarded as a unique case.
It seems to be established that the cyto- and histogenesis of the
nervous tissue is determined in early embryonic life. Various instances
of functional determination of a few parts were discussed in the previous
sections. The question we want to discuss now is whether this morphological determination can simulate the type of behavior which has
been ascribed to "functional specificity."
d. Although the foregoing propositions seem to be more or less supported by experimental data, proposition d may be regarded as an indirect deduction partly from the former propositions, partly from other
considerations—a tentative program rather than a direct statement. It
considers a form-function relationship. An old problem in neurology is
the extent to which anatomy relates to physiology. Several attempts
have already been made to draw functional conclusions from morphological investigations, but we have still not advanced beyond vague
guesses that similar anatomical structures may share similar tasks in
the general activity of the nervous system. This proposition, however,
needs to be a more specific one dealing with the possibility of the organization of intercellular connections in cell assemblies on the basis of the
morphological characteristics of the components.
Although simple conclusions seem to be close at hand, for instance, that
larger receptive fields belong to secondary sensory cells with wider
dendritic arbors, one is still a bit hesitant to attribute an instrumental
role to morphological characters such as cell density, geometry, distances,
convergence, form of contact, bushiness, and diameter in building up the
functional organization in a given cell assembly. In fact, a particularly
attractive suggestion has already been made by Lettvin et al. (1961)
in this context, showing a correspondence between the number of operational groups at the output of the retina and the various anatomical
groups of ganglion cells. Making use of Ramon y CajaPs extensive studies
on the histological structure of the retina, they distinguished five types
of ganglion cells in the frog according to the dendritic arborization pattern. In an earlier work (Maturana et al., 1960), they found five classes
213
determined by intrinsic factors. Different sections of the center retain their
morphological character after being transplanted in heterotopic places
(Detwiler, 1943; Straznicky, 1963), or following complete isolation from
the rest of the nervous system (Hamburger, 1946; Bueker, 1943; Reddick,
1945; B. S. Wenger, 1951). This demonstrates a self-differentiation capacity of the parts. The early belief that the adult amphibian nervous
system is capable of reconstituting extirpated regions of the brain or cord
has not gained support (Lorente de No, 1921; Piatt, 1955). The remarkable capacity of the retinal pigment epithelium to regenerate regionally
specific neural retina (Stone, 1950) may be regarded as a unique case.
It seems to be established that the cyto- and histogenesis of the
nervous tissue is determined in early embryonic life. Various instances
of functional determination of a few parts were discussed in the previous
sections. The question we want to discuss now is whether this morphological determination can simulate the type of behavior which has
been ascribed to "functional specificity."
d. Although the foregoing propositions seem to be more or less supported by experimental data, proposition d may be regarded as an indirect deduction partly from the former propositions, partly from other
considerations—a tentative program rather than a direct statement. It
considers a form-function relationship. An old problem in neurology is
the extent to which anatomy relates to physiology. Several attempts
have already been made to draw functional conclusions from morphological investigations, but we have still not advanced beyond vague
guesses that similar anatomical structures may share similar tasks in
the general activity of the nervous system. This proposition, however,
needs to be a more specific one dealing with the possibility of the organization of intercellular connections in cell assemblies on the basis of the
morphological characteristics of the components.
Although simple conclusions seem to be close at hand, for instance, that
larger receptive fields belong to secondary sensory cells with wider
dendritic arbors, one is still a bit hesitant to attribute an instrumental
role to morphological characters such as cell density, geometry, distances,
convergence, form of contact, bushiness, and diameter in building up the
functional organization in a given cell assembly. In fact, a particularly
attractive suggestion has already been made by Lettvin et al. (1961)
in this context, showing a correspondence between the number of operational groups at the output of the retina and the various anatomical
groups of ganglion cells. Making use of Ramon y CajaPs extensive studies
on the histological structure of the retina, they distinguished five types
of ganglion cells in the frog according to the dendritic arborization pattern. In an earlier work (Maturana et al., 1960), they found five classes