114
A. M. PERKS
makes it difficult to be certain which characteristics of their neurohypophyses are truly primitive, and which are developments peculiar to
each group. Also, it suggests that they should be given separate
consideration.
1. MYXINIFORMES
Despite the difEculties pointed out above, a number of workers have
concluded that the Myxiniformes possess the most primitive pituitary of
any living vertebrate (see Gorbman, 1965). However, in his early studies,
de Beer (1926) had suggested that their pituitaries might be degenerate;
despite this, he was impressed by the development of the infundibular
process and felt that it was better formed than in the lampreys. Unlike
the lamprey, the infundibulum of the hagfish consists of a hollow, flattened sac, 1 mm long, and attached to the brain by a narrow stalk
(Jansen, 1930). Early accounts suggested that this process was poorly
differentiated ( Retzius, 1895; Jansen, 1930; Herlant, 1954) , but recent
work has shown that all the components of a neurohypophysial system
are present in both Myxine glutinosa and in Polistotrema stouti (Olsson,
1959; Honma, 1960; Adam, 1963a; Gorbman, 1965). The following general account of the hagfish neurohypophysis is based on the work of these
recent authors.
In the hagfish, the paired preoptic nuclei consist of ill-defined clusters
of cell bodies, located dorsally to a poorly developed optic tract (Fig. 1).
The more ventral cells (“parvocellularis”) lie close to the preoptic recess
and do not appear to contain neurosecretion. In contrast, the dorsal cells
( “magnocellularis”) contain small quantities of fine, perinuclear neurosecretory granules which stain with Astra blue, but which are not detectable by the classic chrome-hematoxylin-phloxin stain of Gomori (Olsson,
1959; Adam, 1963a). Unlike neurosecretory cells in many other lower
vertebrates, the preoptic cells of the hagfish do not send dendrites to the
cerebral ventricle; however, it should be pointed out that the ventricles
are greatly reduced in these species (Adam, 1963b). Neurosecretion may
leave the nucleus by two possible pathways. The first is probably unique
to the Myxiniformes: It is possible that neurosecretory material may pass
directly from the nucleus into a portal system which leads to the neural
lobe of the pituitary. This is suggested by the presence of accumulations
of secretion between and within the capillary walls of a vascular plexus
which lies directly beneath the preoptic nucleus; this plexus has been
seen to drain to the neural lobe and pars nervosa (Gorbman et al., 1963).
The second pathway is well established in many vertebrates; neurosecretion may leave the nucleus along the extremely thin processes which make
A. M. PERKS
makes it difficult to be certain which characteristics of their neurohypophyses are truly primitive, and which are developments peculiar to
each group. Also, it suggests that they should be given separate
consideration.
1. MYXINIFORMES
Despite the difEculties pointed out above, a number of workers have
concluded that the Myxiniformes possess the most primitive pituitary of
any living vertebrate (see Gorbman, 1965). However, in his early studies,
de Beer (1926) had suggested that their pituitaries might be degenerate;
despite this, he was impressed by the development of the infundibular
process and felt that it was better formed than in the lampreys. Unlike
the lamprey, the infundibulum of the hagfish consists of a hollow, flattened sac, 1 mm long, and attached to the brain by a narrow stalk
(Jansen, 1930). Early accounts suggested that this process was poorly
differentiated ( Retzius, 1895; Jansen, 1930; Herlant, 1954) , but recent
work has shown that all the components of a neurohypophysial system
are present in both Myxine glutinosa and in Polistotrema stouti (Olsson,
1959; Honma, 1960; Adam, 1963a; Gorbman, 1965). The following general account of the hagfish neurohypophysis is based on the work of these
recent authors.
In the hagfish, the paired preoptic nuclei consist of ill-defined clusters
of cell bodies, located dorsally to a poorly developed optic tract (Fig. 1).
The more ventral cells (“parvocellularis”) lie close to the preoptic recess
and do not appear to contain neurosecretion. In contrast, the dorsal cells
( “magnocellularis”) contain small quantities of fine, perinuclear neurosecretory granules which stain with Astra blue, but which are not detectable by the classic chrome-hematoxylin-phloxin stain of Gomori (Olsson,
1959; Adam, 1963a). Unlike neurosecretory cells in many other lower
vertebrates, the preoptic cells of the hagfish do not send dendrites to the
cerebral ventricle; however, it should be pointed out that the ventricles
are greatly reduced in these species (Adam, 1963b). Neurosecretion may
leave the nucleus by two possible pathways. The first is probably unique
to the Myxiniformes: It is possible that neurosecretory material may pass
directly from the nucleus into a portal system which leads to the neural
lobe of the pituitary. This is suggested by the presence of accumulations
of secretion between and within the capillary walls of a vascular plexus
which lies directly beneath the preoptic nucleus; this plexus has been
seen to drain to the neural lobe and pars nervosa (Gorbman et al., 1963).
The second pathway is well established in many vertebrates; neurosecretion may leave the nucleus along the extremely thin processes which make