BREVIORA 
MniseiuitM of Comparative Zoology 
Cambridge, Mass. June 17, 1959 Number 111 
ON THE CAUDAL NEUROSECKETORY SYSTEM OF THE 
TELEOST PISH, FVNDIJLVS HETEROCLITUS L. 
By Uno Holmgren ^ 
Zoologifiil Institute, Uppsala, Sweden 
Biological Laboratories, Harvard University 
INTRODT'CTION 
It Avas early observed that a bulb-like structure was associated 
with the terminal portion of the spinal cord of certain fishes 
(AVeber 1S27, Rauber 1877, Arsaky 1813, Ziehen 1903, and 
P'avaro 1925). This outgrowth of the spinal cord of different 
systematical g-roups of teleosts was described by Favaro (1925), 
who studied a large number of species. Because of the histologi- 
cal and morphological resemblance of this organ to the neuro- 
hypophysis, he named it "Ipofisi caudale." 
Secretory cells in the terminal portion of the spinal cord were 
described first by Dahlgren (1914), and later by Speidel (1919, 
1922). Tlie latter also gave a histological description of some 
"irregular glandular cells," sometimes called " Dahlgren 's 
cells," in the terminal portion of the spinal cord of certain 
elasmobranchs and teleosts. Recently Enami (1955 a) investi- 
gated the caudal portion of the spinal cord of the eel, Anguilla 
japonica, with respect to the secretory activity of " Dahlgren 's 
cells" and found that neurosecretory cells were present in the 
terminal portion of the spinal cord. They extended caudally 
from the level of the last sixth or seventh vertebra. These neuro- 
secretory cells, according to Enami, resembled the secretory cells 
of the hypothalamus, and had axons which served as neurosecre- 
1- FflldW (if the t 'oiiniioinvcaltli l'''ini(1. 
2 i;i;eviora No. Ill 
tory patlnvays. The neuTosecretory niaterial was roleased at 
nerve endings, which terminate in a ventral outgroAvth of the 
s})inal cord. The outgrowth served as a storage depot for secre- 
toi-y material. This sti'ucture, which was named "Ipofisi can- 
dale" by P'avaro (1925), was called "Neurohypophysis spinalis'" 
or ■' rrohyjiophysis" by p]nami (1!)55 a). 
These observations were extended by Enami and Imai (1955. 
1956 a and b), with descriptions of the candal neurosecretory 
system of a luimber of freshwater and marine teleosts. Recently 
Sano (1958 a and b) described the " Neurophysis " and the neuro- 
secretory system of the spinal cord of Tinea ruJgaris and a num- 
ber of other teleosts. It has been suggested that the probable 
function of the caudal neurosecretory system would be osmoregu- 
lation (Enami. ]\Iiyashita and Imai 1956, and Holmgren 1!)5S b». 
Terminolofjij. Different names have been suggested for the 
above-mentioned outgrowth of the spinal cord, which servers as 
a storage organ for the neurosecretory material. Because the 
organ may be located either dorsal, ventral, or lateral to the sjiinal 
cord (Favaro 1925, Sano 1958 b), and since the term "neuro- 
hypophysis" has already been used for the "pars nervosa" of 
the pituitary gland, the name "neurohypophysis spinalis" is not 
adequate. In order to stress the fact that the organ is an ont 
growth of the terminal jiortion of the spinal cord the name 
Kroph I/sis spinalis'^ has been ado])ted for this study. This name 
stresses the position of the organ relative to the body, regai-dless 
of its varying position in relation to the spinal cord. 
Material and nufhods. Twenty specimens of the teleost tish. 
Fundulus heteroclitiis L., standard length 4-() cm., caught during 
different times of the year, were used for the anatomical descrip- 
tion. TAvelve end)ryos were also used in an attem])t to determine 
the mode of development of the urophysis. For the anatondcal 
and histological descriptions, the terminal portion of the spinal 
cord, including tli(> adjoining vertebral column, was fixed in 
Bouin's solution to which 1 per cent of CaCl2 had Ix-cn adde;! : 
Zenker's, Carnoy's and Orth's fixing fluids were also used. Tlic 
tails Avere decalcified in 7 ]ier cent IIXOm and aftci'wai'ds treate;! 
in 10 i)er cent Xa^SO-i, and the sections were cnt a1 4-S mici-ons. 
sagittally and transversely, using the usual pai-affin technique. 
1 SusKestc'l liy l»r. (J. I'riillnTL'. Zii(ilo'_'-ical Iiistitnlr. SIcirUlKilin. 
1059 CAUDAL NEUROSECRETORV SYRTEIM 3 
Tlic i'ollowiiifi- staiiiin.u' t('('liiii(|ii('s wei-e used: (ioinori 's liaema- 
toxylin-phloxin method, Ilcideiilmiii's azaii, llalnii's aldehyde 
tuehsin following performie acid oxidation (Ilolmoren 1958 a), 
Mallory's connective tissue stain, Bodian's ]m)taro'ol method, 
Masson's trichrome stain according- to Gomori (1950), periodic 
acid-Schiff (PAS), and alcian blue. 
DE8CKIPT10N 
Location of the urophysis spinalis. The nrophysis in Fundulus 
is observed as a round body, easily visible under the dissecting 
microscope. It lies at the end of the spinal cord, ventral to the 
terminal portion, above the articulation between the last vertebra 
and the hj-pural bone. This location of the urophj^sis (Figs. 1-4) 
is similar to that described for a great number of teleosts by 
Enami and Imai (1956 a and b), and Sano (1958 a and b). 
These authors also found that the nrophysis is generally situated 
ventral to the spinal cord. 
Morphology and histology. According to descriptions by Fa- 
varo (1925), tlie nrophysis consists mainly of modified glial 
tissue. The blood vessels present in the organ enter from the 
connective tissue mantle which surrounds the spinal outgrowth. 
Rauber ( 1877 ) , on the other hand, believed that the nrophysis 
consisted mainly of connective tissue. Enami (1955 a) found 
both connective tissue and modified glial tissue in the organ, and 
stressed the similarity of the nrophysis to the neurohypophysis 
of vertebrates and the sinus gland of Crustacea. 
The nrophysis spinalis of Funduhis heteroclitus is a very con- 
spicuous structure measuring about 0.2 x 0.3 mm., which ap- 
proximately corresponds to the size of the pituitary gland. It is 
covered by a heavily pigmented leptomeninx. The histological 
elements distinguished in the nrophysis are mainly of three kinds. 
In the dorsal region, nerve fibers are found which emanate from 
the neurosecretory cells, situated in the terminal portion of the 
spinal cord (Fig. 1). Orange G positive Herring body-like secre- 
tory droplets are present at the end of these axons (Fig. 3). 
These 'secretory droplets' are probably dilated nerve endings. 
The colloidal masses within the dilated nerve endings stain red 
in Gomori 's chrome haematoxylin-phloxin, thus showing affinity 
4 BREVIORA No. Ill 
for the phloxin component. The staining' reaction with azan is 
blnislired. The nerve tracts are nsnally not clistribntecl all over 
the organ as observed in some other species but merely confined 
to the dorsal region of the urophysis. 
The nrophysis is very vascular with blood vessels distributed 
throughout the organ (Figs. 4, 6). These vessels are of sinusoid 
type with thin walls, supported by argyrophilic connective tissue 
as shown with Bodian's protargol method. In the dorsal region 
of the urophysis the blood vessels frequently seem to surround 
OTTI^ Neurosecretory cells 
,1 mm. ' 
Fig. 1. Sclieniatical picture of the caudal neurosecretory system of the 
teleost. Funduhis licferoclifus L. 
intercellular sjiaces, which are most conspicuous in the area 
where the secretory tracts terminate. Besides the nerve endings, 
the urophysis consists of (n-f/ipapliilic connective tisfiuc and neu- 
roglia. The gli.d cells contain very little cyto])lasm, sometimes 
consisting only of small nuclei and surrounding membranes. The 
differentiation lietween the different types of tissue elements is 
especially well observed after using Heidenhain's stain. The 
distribution of the blood vessels indicates that the secretory ma- 
terial is taken up l)y the blood. 
Transverse sections show that the uro])hysis is surrounded liy 
a sella turcica-like connective tissue capsule (Fig. 4). This ex- 
tension of the leptomeninx is sometimes very heavily invested 
with melanophoi-es. In other teleosts, the urophysis may be at- 
tached to the spinal cord by means of a stalk, which sometimes 
is very conspifuoiis (Eiiami ;md Inuii 1956 a and 1), ?>ano 1958 b). 
1959 CAUDAL NEUROSECRETORY SYSTEM 5 
The nr()i)hy.sis of FiokIhIus lieteroclitus is very closely attached 
to the spinal cord by means of a stalk with a broad base. Most 
of the org-an, however, is separated from the spinal cord. Cross- 
sections of the urophysis (Fig. 4) show a well defined space be- 
tween the dorsal part of the organ and the ependymal parts of 
the spinal cord. Reissner's fiber is still present in this thin ter- 
minal portion of the spinal cord. 
In tlie region where the spinal curd axons i)ass into the nro- 
physis the secretory droplets are especially frequent (Fig. 3). 
The secretory material within the dilated nerve endings is here 
homogeneous and appears colloidal. The staining reaction also 
seems to have changed in a 'basophilic' direction (cf. below). 
Some of the secretory droplets were slightly aldehyde fuchsin- 
positive in the dorsal region of the urophysis, but the neurosecre- 
tory cells stained negatively. 
The neurosecretory cells. The neurosecretory cells are present 
in the terminal portion of the spinal cord, appearing first at the 
level of the sixth vertebra from the caudal end and most concen- 
trated above the third vertebra and caudally (Figs. 1, 5). The 
secretory cells, which contain Xissl substance, are obviously spe- 
cialized nerve cells of multipolar, bipolar, and also unipolar type. 
The neurosecretory cells stain negatively with both Gomori's 
chrome haematoxylin and Halmi's aldehyde fuchsin. Instead, 
they take up the pliloxin component of Gomori's stain and also 
show great affinity for the acid fuchsin of ]\Iallory's stain, thus 
confirming the findings of Enami (1955 a). 
The neurosecretory cells are also stained with the azocarmine 
of Heidenhain's stain thus .showing a marked 'acidophilic' reac- 
tion which may be due to secretory- material in the cytoplasm 
(cf. below). The spinal neurosecretory cells were observed to 
contain scattered small PAS-positive inclusions, but the secretory 
droplets within the urophysis stained negatively. A positive 
reaction would have denoted the presence in the secretion of 
considerable amounts of polysaccharides, mucopolysaccharides, 
glucoproteins or giycolipoids (Lillie ]954). The negative reac- 
tion to PAS agrees with the findings of Sano (1958b). 
The neurosecretory cells have various sizes and shapes. Those 
which are situated closer to the urophysis, are generally smaller. 
The secretory cells are generally distributed all over the spinal 
6 BREVIORA No. m 
cord which in tho tenninal portion does not show the nsual sep- 
aration into white and gray material. The secretory cells were 
found to contain a larjie nucleus with an ii-regular shape, as 
already indicated in the early descriptions by Speidel (1919, 
1922). Tlie size of the neurosecretory cells in FiduIuIks, how- 
ever, was considerably less than the ori<i'inally described spinal 
cells in skates and the eel (Speidel 1919, Enami 1955a). The 
secretory niatei-ial usually contained very fine granules, and was 
generally distributed all through the cytoplasm but sometimes 
was merely confined to the periphery of the cells. The irreg- 
ular nucleus contains one or more nucleoli and fine granular 
chromatin. 
In the sections stained with lleideidiaiu's azan, it was easy to 
distinguish tlie neurosecretory cells, with their 'acidophilic' re- 
action, from the strongly 'basophilic' motor cells. Because of the 
irregular shape of the nuclei, sometimes two or several nuclear 
areas were visible when a cell was cut in a certain plane. In the 
cytoplasm of the secretory cells the 'acidophilic' material was 
observed in various amounts, indicating a cycle in the formation 
of the secretion. The 'acidophilic' reactioii was interpreted as a 
measure of the secretory activity. Nerve cells in general and 
some neurosecretory ones from the hypothalamic system may 
show a 'basophilic' reaction (Seharrer and Scharrer 1954), per- 
haps due to a continuous synthesis of proteins, hormones and 
high amounts of ribonucleic acid. Observations on cells in dif- 
ferent stages of secretion have indicated that the cell in the be- 
ginning of the secretory phase does not show a considerable 
'acidophilic' reaction. As secretion begins the luu-lei of the cells 
begin to enlarge and 'acidophilic' granulated material arises in 
the area around the nucleus. Later, the secretory material is 
often found in the periphery of the cells or distributed all 
through the cytoplasm. It has been suggested that the heavy 
'acidophilic' reaction particularly in the periphery of the neuro- 
secretory cells is due to accumulation in these areas of 'acido- 
philic' material and Nissl-substance (>Sano 1958 b). The ob- 
servations in this study suppoi't the suggestion by Sano (1958 b) 
that the nucleus takes an active part in the formation of the 
secretory material. No selective stain is so far known for the 
secretoi'v material. 
19.')fl CAinAI, XKIROSECRETOKY SVSTKM 7 
The secretury material of tin- lUTve cells is traiisi)orle(l in the 
axons whieh serve as netirosecretory pathways and is stored in 
dilated nerve endiniis jiiid released in the dorsal region of the 
iirophysis where the axons terminate. Observations were made 
on the staining reaction of the secretory droplets to acid alcian 
bine stain. The observed negative reaction wonld mean that the 
secretion does not contain disnltide groups. These observations 
agree with those by Sano (1958 b), who found that the secretion 
in the caudal neurosecretory system of Tinea vulgaris reacted 
negatively for the astra blue and alcian blue stains. 
In the limited material examined (20 species), obtained dur- 
ing fall and wdnter, no pronounced seasonal differences were 
observed in secretory activity. 
Development of the urophysis spinalis. The caudal neuro- 
secretory cells and the urophysis spinalis were not developed 
in the stages immediately after hatching, length 15-20 mm. (Fig. 
7). The hypothalamus-pituitary system was apparently actively 
secreting at this stage. These observations thus confirm the find- 
ings by Favaro (1925) and Sano (1958 b) that the urophysis 
spinalis and the caudal neurosecretory" system develop late dur- 
ing the ontogeny. Sano (1958 b) noted that the urophysis spinalis 
was still not developed in Salmo fario at a size of 2.5-3 cm. In 
Anguilla of 6 cm. length, the caudal neurosecretory system had 
begun to function. 
-e " 
DISCUSSION 
Many investigators have pointed out the similarity in struc- 
ture between the posterior pituitary and the urophysis spinalis, 
and this analogy has also been stressed in the names suggested, 
e.g. "Ipofisi eaudale" (Favaro 1925), ''Neurohypophysis spin- 
alis" (Enami 1956), and "Neurophysis spinalis caudalis" (Sano 
1958 a, and Sano and Hartmann 1958). This similarity in struc- 
ture and organization of the urophyseal system compared to the 
hypothalamic system was emphasized in this study by the fact 
that the secretory material in the urophysis was stored in dilated 
nerve endings. 
As alread}' mentioned al)ove, there is no selective stain for the 
secretory material in the caudal neurosecretory system, although 
the dilated nerve endings show attinitv for the orange G com- 
8 BREVIORA No. Ill 
poiient of various stains. There is. however, reason to believe that 
the 'acidophilic' reaction of the cells is due to the presence of 
secretion. The terms 'acidophilic' and 'basophilic' as used in 
this study are relative, depending on fixation, stainino- and other 
treatment. The two terms have been used to indicate that the 
number of eitlier acid carboxyl or basic amino groups of the pro- 
tein molecule may be predominant (Romeis 1948). These terms 
may be useful in this case for descriptive purposes until a more 
selective stain for tlie caudal neurosecretory system has been 
developed. 
There are reasons to believe that the neurosecretory cells in the 
teleost s])iiial eoi'd, although they diffei- in size from the cells 
originally described by Dahlgren (1!)14) and Speidel (1919), 
are identical and homologous with those in the skates, in spite of 
the fact that Speidel (1922) did not find the corresponding cells 
in the spinal cord of Fundnli<s. The neurosecretory cells of 
Fundidtis do not differ greatly from unmodified nerve cells, ex- 
cept for their staining reactions (cf. above). They are far less 
conspicuous and appear less "glandular" than the secretory cells 
of the skates and the eel and can therefore be readily overlooked. 
Sano (1958 b) found that the neurosecretory material inside 
the axons could not l)e a pathological product of the cells or of 
the myelin sheath but must be considered as a true neurosecre- 
tion, as earlier indicated by Enami (1955 a, 1956). This study 
can further support these opinions by showing the cons])icuous. 
Herring body-like, secretory droplets which are formed at the 
nerve endings in the dorsal region of the urophysis spinalis 
(Fig. 3). These secretory droplets are, according to this study, 
the dilated nerve endings, which serve to store the secretory 
material. The secretory material is thus generally not stored in 
the urophyseal tissues (cf. Enami 1955 a). 
The origin of the secretory granules inside the neurosecretory 
cells has been extensively dealt with by Scharrer (1934), Palay 
(1943), Hild (1950), Scharrer and Scharrer (1954), Enami 
(1955 b), and others. According to these descriptions, the foi-- 
mation of the secretion is not fully understood. It has been ob- 
served that granules also are present in the inicleus of certain 
cells of the nucleus tubcralis lateralis of teleosts (Scharrer 1934, 
Stahl and Seite 1955. Enami 1955 b. Ortiuann 1956. Stahl 1957). 
1 !).")!) CAUDAL NEUROSECRETORY SYSTEM 9 
Acc()r(liii<i- to Stalil and Seite (1955), the iiitrHinu'lcar granules 
l)i-()])al)ly cannot be eonneeted with neurosecretion. Enanii (1955 
h), on the other hand, considered that tlie 'acidophilic' <iTaniiles 
iu the nucleus entered the cytoplasm through the nuclear mem- 
brane into the area around the nucleus or were emitted into 
neurites. The direct role of the nucleus in the secretory process 
has not yet been clearly demonstrated. As far as the caudal 
neurosecretory system is concerned, 8ano (1958 b) believes that 
tlu' nucleus vciy likely takes an important part in the formation 
of the secretion. The opinion of Scharrer and Scharrer (1954) 
and Enami (1955) that the secretory material may be formed out 
of the Nissl substance or at the expense of the latter, could not 
be confirmed in this study. As mentioned above, it was observed 
that the secretory material seemed to originate from the area 
around the nucleus, which was interpreted to mean that the 
nucleus plays an important role in the formation of the secretion. 
It was previously noted that the secretion in the neurosecretory 
cells reacted negatively with Gomori's chrome haeniatoxylin and 
Ilalmi's aldehyde fuchsin. The secretory droplets (colloid 
masses) in the dorsal region of the urophysis (Fig. 3), on the 
other hand, showed some affinity for the aldehyde fuchsin. This 
difference in staining reaction with aldehyde fuchsin in the secre- 
tory cells as compared to the urophyseal colloid may indicate that 
physical or chemical changes have occurred in the secretory ma- 
terial in association with its storage in the nerve endings. A 
stronger reaction to Gomori's stain is also generally observed in 
the hypothalamus-pituitary system for the Herring bodies com- 
pared with tlie neurosecretory cells or the tracts. In the hypo- 
thalamus-]) ituitary system, the presence of a stainable carrier 
substance or prosthetic group, has been postulated (Schiebler 
1951). If biochemical changes do occur at the release of the 
neurosecretory material at the nerve endings, it would affect the 
response to certain dyes. 
The negative reaction of the caudal neurosecretory cells and 
the secretory material in the urophysis to the periodic-acid -Schiff 
stain is in contrast with the positive reaction in the hypothalamus- 
pituitary system. Schiel)ler (1951) found that the neurosecretory 
cells of the hypothalamus of the teleost Eso.r were strongly peri- 
odic-Schiff positive. 
10 BREVIORA No. Ill 
The studies by Eiiami (lI)5o a) and Saiio (1958 a and b) have 
demonstrated the presence in fishes of a caudal neurosecretory 
system whieli morpholoo'ically is orp-anized in nnich the same 
manner as the hypothalamus-])ituitary system in vertebrates and 
the intercerebralis-corpus cardiaeum system in insects, all show- 
ing- characteristic synthesizing, transporting-, and release-storage 
elements. Such a neurosecretory system is, according to this 
study, present and actively secreting in the teleost fish, Fundulus 
heteroclitus L. 
ACKNOWLEDGEMENTS 
Dr. Gunnar Pridlierg, Zoological Institute, Stockholm, has read 
the manuscript and oft'ered valuable suggestions. The help of 
Dr. A. B. Dawson, Biological Laboratories, Harvard University, 
in preparing the manuscript is gratefully acknowledged. 
SUMMARY 
An examination of the terminal portion of the spinal cord of 
the teleost Fioxluhis heteroclitus L. has confirmed the presence 
of a caudal neurosecretory system. The neurosecretory cells are 
distributed from and posterior to the sixth vertebra and are more 
concentrated above the third vertebra from the caudal end of the 
spine. Histologically, these neurosecretory cells are modified 
nerve cells of multi- ])i-. and unipolar type. They show affinity 
to 'acidophilic' stains such as acid fuchsin, phloxin, and azocar- 
inine. In contrast with the neurosecretory cells of the hyjiothala- 
nnis, these s[)inal cells are Gomori-iiegative and show no reaction 
for the alcian blue stain, the latter staining reaction indicating 
the absence in the secretion of sulfonate groups. The secretory 
matefial, wliich also shows negative periodic-acid-Schiff reaction, 
is transported from the neurosecretory cells in axons serving as 
secretory ]iathways, and the neurosecretion is stored in dilated 
nerve endings wliich terminate in a conspicuous organ, the U)'o- 
plnjsis spinalis (neurohypophysis si)iualis caudalis, neurophysis, 
ur()hyi)oi)hysis), which is a ventral outgrowtli of the spinal cord. 
At the iici-ve endings. Herring l)()dy-like secretory droplets ai-e 
formed. No seasonal changes in the secretory activity of the 
caudal ncMirosecretory system were observed. Tlie uro]ihysis 
shows histologically the same structure as the neurohypophysis 
1959 CAUDAL NEUROSECRETORY SYSTEM 11 
of the i)ituitary ^'laiid, containing nerve eiulinj^s, modified <rlial 
tissue and blood vessels. 
The nrophysis and the caudal neurosecretory system must de- 
velop rather late durin<; the ontop:eny since it is not present in 
end)ryos immediately after hatehin*^, at which time the hypo- 
thalamie-|)ituitary system is actively secreting;. 
REFERENCES 
.\ksakv, a. 
1813. Uc pisum fciobro ct nu'dulln spinali. Disseitatio inauguralis. 
Halae, p. 6. 
Hahlgren, U. 
1914. The electric motor nerve centers in the skates (Ra.iidae). Science, 
40:802-863. 
EXAMI, M. 
1955a. Studies in neurosecretion. II. Caudal neurosecretory system in 
the eel {AnguiUa juponica). Gunma Journ. Med. Sci., 4(1): 
23-36. 
1955b. Nuclear secretion in the cells of the preoptic nucleus in the eel 
{AnguiUa japonica). Endocrin. Jap., 2:33-40. 
19o(). Studies in neurosecretion. VIII. Changes in the caudal neuio- 
secretory system in the loach (Misgurus aiiguillicaudatiu'i) in 
response to osmotic stimuli. Proc. Jap. Acad., 32(3 ) :197-200. 
Ena-mi, M. and K. Imai. 
1955. Studies in neurosecretion. V. Caudal neurosecretory system in 
several freshwater teleosts. Endocrinologica Jap., 2(2):105-116. 
195(5n. Studies in neurosecretion. VI. Xeurohypophysis-like organize 
tion near the caudal extremity of the spinal cord in several 
estuarine species of teleosts. Proc. Jap. Acad., 32:197-200. 
1956b. Studies in neurosecretion. VII. Further observations on the 
caudal neurosecretory system and the neurohypophysis spinalis 
(urohypophysis) in marine teleosts. Proc. Jap. Acad., 32(8): 
633-38. 
En'ami, M., S. Miyashita and K. Imai. 
1956. Studies in neurosecretion. IX. Possibility of occurrence of a 
sodium-regulating hormone in the caudal neurosecretory system 
in teleosts. Endocrinologies Jap., 3(4) :280-90. 
Favaro, G. 
1925. Contributi alio studio morphologico dell' ipofisi caudale dei 
tcleostei. Mem. K. Accad. Naz. Lincei, 6(l):29-72. 
12 BREVIORA No. Ill 
(iOAIOKI, G. 
1950. A rapid one-step tiiclironie stain. Amer. Jonr. Clin. Path., 20: 
661-664. 
HiLi), W. 
1950. Zur Frage der Neurosecretion ini Zwischenhirns der Sehleie 
(Tinea vulgaris) iind ihrer Beziehung- zur Xeurohypophyse. 
Zeitschr. Zellforsth., 35:33-40. 
Halmi, S. N. 
1952. Differentiation of two types of basophils in the adenohypophysis 
of the rat and the mouse. Stain. Tech., 27(l):61-64. 
Holmgren, U. 
19o8a. Secretory material in the pineal body as shown by aldehyde- 
fuchsin following performic acid oxidation. Stain. Tech., 33(3) : 
147-149. 
19581). On the caudal neurosecretory system of tlie teleost fish, Fun- 
liitiiLs hffcrnriitii.s. (Abstract.) Anat. Rec. 132:454. 
Lll,LIE, li. U. 
1954. Histopathologic technic and practical histoclicmistry. Blakiston 
Co., New Yoik, Toionto. 
Ortmax.v, R. 
1956. Ilistochemische Untersucluingen an den Kerneinschlusskorperu 
der Ganglienzellen ini Nucleus praeopticus dcs Karpfen (Cypri- 
ntis carpio L.). Zeitschr. Anat. Entw. Ges., 119:485-499. 
I'ALAV, S. L. 
1943. Neurosecretion. A'. The origin of neurosecretory granules from 
the nuclei of nerve cells in fishes. Jour. Comp. Neurol., 79: 
247-275. 
IvArBKR, A. 
1877. Die letztcu spinalen Nerven und (ianglien. Morph. Jahrl)., 3: 
603-624. 
Ko>tEis, B. 
1948. Mikroskopische Teehnik. Munich 15th ed. 
Saxo, Y. 
1958a. ITeber die Xeuiophysis (sog. Kaii(lalliy|io])liyse, "UrohyjJO- 
physe") lies Teleostieis, Tinea nih/ari.s. Zeitschi-. Zellforsch., 
47:481-497. 
19.")8li. W'eitere Unteisuchungen iil)er den Feinliau der Neuroi>liysis 
spinalis caudalis. Zeitsdir. Zellforsch., 48:236-260. 
If).')!) CAUDAL NEUROSECRETORV SYSTEM 1 :> 
Sano, V. ;uid F. Haktmann. 
1!)58. Zur voigleiehondcii Ilistolojiic \()ii Neuropil^ sis spiimlis cini- 
il;ilis iiiul Xciiroh.vpdpli.vsis. Zcitsclir. Zcllt'orscli., 48:r);!S-r)47. 
SCHARKKK. K. 
1!>;>4. Uehor den r'oitoilisjunf>t'n des ZeUkciiis .-m sckictorisclicn 
Voifjiin^en in Ni'ivcn/.cUcii. P^niikf. Zcitsclir. l':itli., 47:14IMm. 
SrnAKRER, E. and B. Scharrkr. 
1954. Xi'nrosecTction. In Ilnndli. inikr. Anat. Mciisch., 4(r) ) : !»r):M050. 
SCHIEBLKR, T. H. 
19;')!. Zur Histoflu'niie dos nonrosecrotoriscdicn hypotlinlainiscdi-neuro- 
hypophysaren Systems. Acta Anat., 13:233-tir)r). 
Speidel, C. C. 
1919. Gland-cells of internal secictiou in the spinal cord of skates. 
Carnegie Inst. Washington, Pul)l. 1181. 
W^'l'l. Further comparative studies in other fishes of cells homologous 
to the large irregular glandular cells in the spinal cord of the 
skates. Jour. Comp. Neur., 34:303-317. 
STAilL, A. 
1957. Kecherches sur les elaliorations cellulaires et la neuiosecretion 
dans I'eneephale des poissons teleosteens. Acta Anat. (Basel), 
31(suppl.) :28. 
Wtahl, A. and R. Seite. 
1055. Sur la presence d'une elaboration figuree d'origine nucleaire 
dans les cellules neurosecretoires du noyau preoptique chez les 
poissons teleosteens. C.R. Soc. Biol. (Paris), 149:382. 
WEiiEi;, 10. II. 
1827. Knoten uud uupaaieu Faden iiiit dem sich das Iliickenuiaik liei 
einigen Fisclien endigt, namentlich l>eim Ci/priiiu.s cdrpid. Mcch- 
el's Arch. Anat. u. Physiol., 2:31(i-317. 
Ziehen, Th. 
1903. rVutralnervensystem, 1 Aid. Jena. Gustav Fischer.