168
A. M. PERKS
vesicles may be clustered in the center of the fibers or they may lie along
an electron-dense area next to the neural membrane. In some cases, the
neural membrane is thickened at this point. These endings, which are
apparently synaptic terminals, also contain elementary vesicles; and in
Anguilla anguilla these structures appear to break down, lose their electron density, and perhaps release small particles into their surroundings
(Gadus morrhua; Lederis, 1962; Lebistes reticulatus; Follenius and Porte,
1962; Anguillu anguilla; Knowles and Vollrath, 1965a, 1966a).
The terminals of the neurosecretory fibers can be found in four main
places: They occur on pituicytes, within the pars intermedia, on the outer
intervascular channel, or around capillaries. Knowles and Vollrath
(1965a,b, 1966a) have given a clear description of synapses between
neurosecretory fibers and pituicytes in both Anguilla anguilla and Conger
conger. The number of synapses between pituicytes and Type A, terminals appeared to increase when specimens of Anguilla anguilla were
transferred to seawater, while a greater number of synapses with Type A,
fibers were found in eels which had been placed on an illuminated white
background. This suggestion of the formation of “transient” synapses between nerve fibers and pituicytes may prove to be of considerable
importance.
The extent to which neurosecretory terminals penetrate directly into
the pars intermedia tissue varies with different species, and it may well
depend on the size of the fenestrations in the intervascular channel, which
covers the surface of the digitations of the pars nervosa and corresponds
to the ‘%basement membrane” of earlier authors. In Conger conger there
are only small areas of intervascular space which divide the pars nervosa
from the pars intermedia, and one finds considerable intermingling of
nerve fibers and intermedia cells, with the formation of synaptic connections (Knowles and Vollrath, 1966a). Similarly, in Gadus and Phoxinus
there is little dividing structure between the two regions, and direct
neuroglandular contacts have been reported ( Follenius, 1965). In
Lebistes reticulatus it is probable that the dividing structures are partly
developed, and a minority of the neurosecretory axons end against pars
intermcdia cells ( Follenius, 1965; Follenius and Porte, 1962). In Salmo
irideus the barrier is probably better formed, and only a few fibers pass
through it, while in Perca fluuiatilis, where the dividing structures are
well developed, nerve terminations in the pars intermedia are rare
(Follenius, 1965; Legait and Legait, 1957; Follenius and Porte, 1962). In
Anguilla anguilla a few neurosecretory axons of type A, pass through gaps
in the outer barrier (intervascular channel), but no synaptic terminals
have been found on intermedia cells (Knowles and Vollrath, 1966a).
In Anguilla anguilla, by far the majority of the neurosecretory
A. M. PERKS
vesicles may be clustered in the center of the fibers or they may lie along
an electron-dense area next to the neural membrane. In some cases, the
neural membrane is thickened at this point. These endings, which are
apparently synaptic terminals, also contain elementary vesicles; and in
Anguilla anguilla these structures appear to break down, lose their electron density, and perhaps release small particles into their surroundings
(Gadus morrhua; Lederis, 1962; Lebistes reticulatus; Follenius and Porte,
1962; Anguillu anguilla; Knowles and Vollrath, 1965a, 1966a).
The terminals of the neurosecretory fibers can be found in four main
places: They occur on pituicytes, within the pars intermedia, on the outer
intervascular channel, or around capillaries. Knowles and Vollrath
(1965a,b, 1966a) have given a clear description of synapses between
neurosecretory fibers and pituicytes in both Anguilla anguilla and Conger
conger. The number of synapses between pituicytes and Type A, terminals appeared to increase when specimens of Anguilla anguilla were
transferred to seawater, while a greater number of synapses with Type A,
fibers were found in eels which had been placed on an illuminated white
background. This suggestion of the formation of “transient” synapses between nerve fibers and pituicytes may prove to be of considerable
importance.
The extent to which neurosecretory terminals penetrate directly into
the pars intermedia tissue varies with different species, and it may well
depend on the size of the fenestrations in the intervascular channel, which
covers the surface of the digitations of the pars nervosa and corresponds
to the ‘%basement membrane” of earlier authors. In Conger conger there
are only small areas of intervascular space which divide the pars nervosa
from the pars intermedia, and one finds considerable intermingling of
nerve fibers and intermedia cells, with the formation of synaptic connections (Knowles and Vollrath, 1966a). Similarly, in Gadus and Phoxinus
there is little dividing structure between the two regions, and direct
neuroglandular contacts have been reported ( Follenius, 1965). In
Lebistes reticulatus it is probable that the dividing structures are partly
developed, and a minority of the neurosecretory axons end against pars
intermcdia cells ( Follenius, 1965; Follenius and Porte, 1962). In Salmo
irideus the barrier is probably better formed, and only a few fibers pass
through it, while in Perca fluuiatilis, where the dividing structures are
well developed, nerve terminations in the pars intermedia are rare
(Follenius, 1965; Legait and Legait, 1957; Follenius and Porte, 1962). In
Anguilla anguilla a few neurosecretory axons of type A, pass through gaps
in the outer barrier (intervascular channel), but no synaptic terminals
have been found on intermedia cells (Knowles and Vollrath, 1966a).
In Anguilla anguilla, by far the majority of the neurosecretory