2. THE NEUROHYPOPHYSIS
165
stained tracts may have originated partly in the nucleus lateralis tuberis,
and recently this suggestion has received support from the electron microscope studies of Knowles and Vollrath, ( 1965c, 1966b). Knowles and Vollrath found that in A. anguilla, the greater proportion of the nerve fibers
which invaded the pro- and meso-adenohypophysis did not stain by
classic neurosecretory methods; but, nevertheless, they contained irregular
neurosecretory vesicles, with central electron-dense granules, about 700 A
in diameter (Type B). Similar structures were found in the nucleus
lateralis tuberis. The neural processes within the rostral regions of the
adenohypophysis contained few pituicytes and no central canals to link
them to the ventricles. They appeared to discharge neurosecretory products into the perivascular and intervascular spaces which permeated the
rostral adenohypophysis. Knowles and Vollrath ( 1966b ) suggested that
these spaces represented the hypophysial portal system, while the fibers
themselves corresponded to an internal median eminence. This is in
agreement with the suggestions of other authors, who felt that the rostral
region of the pars nervosa, as recognized by Diepen (1954) and by
Leray and Stahl ( 1961), should be regarded as the functional equivalent
of a median eminence ( Follenius, 1965; Klein, 1967). However, it must be
remembered that Knowles and Vollrath (1965c, 1966b) also recognized
that the rostral neural digitations contained a smaller number of axons
with classic, Type A elementary vesicles, 1400 A in diameter, and similar
to those of the preoptic nucleus. The classic neurosecretory materials were
rare in young specimens but became more abundant during maturity.
The great majority of the classic, Gomori-positive, nonmyelinated
neurosecretory fibers ( Type A ) of the preoptico-hypophysial tract appear
to enter the fingerlike processes of the caudal region of the pars nervosa.
These penetrate extensively and intimately into the pars intermedia
( Anguilla vulgaris; Stutinsky, 1953: A. anguilla Jasinsky, 1961; Leatherland et al., 1966; Knowles and Vollrath, 1966a: Mugil cephalus; Leray and
Stahl, 1961 : Lucioperca lucioperca; Jasinski, 1962: Gasterosteus acukatus,
Perca fluviatilis; Dodd and Kerr, 1963; Hilsa ilisha, Porichth ys notatus,
Sathyanesan, 1963, 1965b: Salmo salar; Klein, 1967: Salvelinus fontinalis,
Henderson, 1969). The caudal neural processes contain a number
of different structures. A few myelinated nerve fibers have been
seen in Perca fluviatilis and Anguilla anguilla (Follenius and Porte,
1962; Knowles and Vollrath, 1966a). Ganglion cells occur in Salmo
trutta, Gadus sp., and Mystus vittatus (Pickford and Atz, 1957; Singh
et al., 1962). Pituicytes has been recognized in Salmo irideus, Cymutogmter aggregata, Gasterosteus aculeatus, Lebistes reticulatus, and Anguilla anguilla; in the last two species they appear to be localized
mainly in the center of the neural digitations, where their processes isolate
165
stained tracts may have originated partly in the nucleus lateralis tuberis,
and recently this suggestion has received support from the electron microscope studies of Knowles and Vollrath, ( 1965c, 1966b). Knowles and Vollrath found that in A. anguilla, the greater proportion of the nerve fibers
which invaded the pro- and meso-adenohypophysis did not stain by
classic neurosecretory methods; but, nevertheless, they contained irregular
neurosecretory vesicles, with central electron-dense granules, about 700 A
in diameter (Type B). Similar structures were found in the nucleus
lateralis tuberis. The neural processes within the rostral regions of the
adenohypophysis contained few pituicytes and no central canals to link
them to the ventricles. They appeared to discharge neurosecretory products into the perivascular and intervascular spaces which permeated the
rostral adenohypophysis. Knowles and Vollrath ( 1966b ) suggested that
these spaces represented the hypophysial portal system, while the fibers
themselves corresponded to an internal median eminence. This is in
agreement with the suggestions of other authors, who felt that the rostral
region of the pars nervosa, as recognized by Diepen (1954) and by
Leray and Stahl ( 1961), should be regarded as the functional equivalent
of a median eminence ( Follenius, 1965; Klein, 1967). However, it must be
remembered that Knowles and Vollrath (1965c, 1966b) also recognized
that the rostral neural digitations contained a smaller number of axons
with classic, Type A elementary vesicles, 1400 A in diameter, and similar
to those of the preoptic nucleus. The classic neurosecretory materials were
rare in young specimens but became more abundant during maturity.
The great majority of the classic, Gomori-positive, nonmyelinated
neurosecretory fibers ( Type A ) of the preoptico-hypophysial tract appear
to enter the fingerlike processes of the caudal region of the pars nervosa.
These penetrate extensively and intimately into the pars intermedia
( Anguilla vulgaris; Stutinsky, 1953: A. anguilla Jasinsky, 1961; Leatherland et al., 1966; Knowles and Vollrath, 1966a: Mugil cephalus; Leray and
Stahl, 1961 : Lucioperca lucioperca; Jasinski, 1962: Gasterosteus acukatus,
Perca fluviatilis; Dodd and Kerr, 1963; Hilsa ilisha, Porichth ys notatus,
Sathyanesan, 1963, 1965b: Salmo salar; Klein, 1967: Salvelinus fontinalis,
Henderson, 1969). The caudal neural processes contain a number
of different structures. A few myelinated nerve fibers have been
seen in Perca fluviatilis and Anguilla anguilla (Follenius and Porte,
1962; Knowles and Vollrath, 1966a). Ganglion cells occur in Salmo
trutta, Gadus sp., and Mystus vittatus (Pickford and Atz, 1957; Singh
et al., 1962). Pituicytes has been recognized in Salmo irideus, Cymutogmter aggregata, Gasterosteus aculeatus, Lebistes reticulatus, and Anguilla anguilla; in the last two species they appear to be localized
mainly in the center of the neural digitations, where their processes isolate