The visual projection patterns of retinal efferents were studied in larval lchthyophis kohtaoensis by means of anterogradely trans- ported HRP. Our results show in all larvae a projection contralateral to a thalamic terminal field, a pretectal terminal field, and a basal optic neuropil, but only a sparse innervation of the contralateral tectum. In addition, all larvae possess an uncrossed projection to a thalamic and a pretectal terminal field. The fibers are bilaterally almost confined to the medial optic tract with only a few fibers run- ning in the marginal and basal optic tract. The ipsilateral and contralateral tracts and terminal fields seem to enlarge during larval life. Comparison with other amphibian orders reveals that larval Ichthyophis are unique in that they develop the medial optic tract and the related thalamic and pretectal terminal fields very early in larval life. In addition they possess only a very sparse tectal projection, though it is the largest projection in larval urodeles and anurans. This suggests a selective phylogenetic loss of those ganglion cells or collaterals which project mainly to the tectum in other amphibian orders and a change in the ontogenetic program leading to an earlier development of the medial optic tract in Ichthyophis as compared to urodeles and anurans.
INTRODUCTION
It has long been bel ieved that non-mammal ian ver-
tebrates possess comple te ly crossed visual projec-
tions, whereas only mammals have deve loped an ad-
ditional uncrossed (ipsilateral) pro jec t ion 1. In recent
years, this view has been chal lenged by findings of bi- lateral project ions in lampreys 33,34,68, selachians30, 48,
polypter i formes 51, ray-f inned fishes10.11.49.52. 60, dip-
noans 47, amphibians t5.39.67, and some sauropsidi-
ans 2,21,57,64. These discoveries have led to the as-
sumption that bi lateral ret inal projec t ions may repre-
sent the primit ive or p les iomorphic state among ver-
tebrates 12. However , within a group of ver tebra tes
such as the agnathans and the lungfish, those species
with a more reduced visual system lack the uncrossed
component of the visual projection37. 46. On the o ther
hand, among teleosts both bi la teral and complete ly
crossed retinal projec t ions have been repor ted 52,
from which it was concluded that the ipsi lateral pro-
jection seems to d isappear within the te leos tean lineo
age 49,52. Thus, both reductive phenomena of the
whole visual system (lungfish,agnathans) and elab-
orat ion of the visual system (teleosts, sauropsidians)
seem to be l inked with the reduct ion of the ipsilateral
retinal project ion. Moreover , among reptiles, reduc-
tion of the visual system seems to be l inked with in-
creased bilateral projection21,~5. A m o n g mammals ,
reduction of the ipsilateral projec t ion coincides with
albinism 19.31. The same holds true for catfish 10 but
not for salamanders20. In the lat ter case, however,
the animals studied have been artificially produced
albinos zT. During deve lopment of mammals and
birds a4,4°.65 a transient ipsilateral project ion becomes
reduced or is 10st14,28,5 o. Enucleat ion of young or lar-
val animals enhances the ipsilateral project ion in anurans, chickens and mammals26, 29,32,63,66, presum-
ably by reducing some sort of competition14. 40. Thus
the ipsilateral projec t ion poses a number of as yet un-
solved questions about its phylogeny as well as onto- geny.
Quest ions with respect to the ipsilateral project ion
are especially obvious in amphibians: different devel-
opmenta l pat terns have been repor ted for the ipsilat-
eral project ions of anurans and urodeles. Among
frogs the deve lopment of the ipsilateral project ion is
Correspondence: B. Fritzsch, University of Bielefeld, Faculty of Biology, Postbox 8642, 4800 Bielefeld 1, F.R.G.
induced by thyroxine changes during metamorpho- sis >. Among salamanders the ipsilateral retinal pro- jection starts to develop prior to metamorphosis 5~, but increases in extent during and after metamorpho-
sis, presumably under the influence of thyroxine~,:. In adult anurans and urodeles, the size of the ipsilateral projection coincides roughly with binocularity3S.39-5<
This seems not to be true for the third amphibian or-
der, the gymnophionans. Here, the adult specimens show almost no binocular visual field but have an ex- tensive bilateral retinal projection 4.24. Likewise, al- most identical bilateral retinal projections, but no ob- vious binocular visual field is found in adult lam- preys 3<~s and in blind snakes 2t.
The present study was designed to investigate the
following questions about visual projections of larval Ichthyophis: (1) does the development of an ipsilat- eral projection in Ichthyophis coincide with meta-
morphosis, as in anurans, or is this projection already formed in the larvae, as is the case in urodeles?; (2) is the development of ipsilateral projections in lchthyo- phis a gradual process, as in urodeles or does it in- volve excessive embryonic projections with subse- quent reduction, as in birds and mammalslT?; and (3) is there any connection in Ichthyophis between bin- ocular vision and the development of the ipsilateral projection, as there is in both anurans and urodeles?
MATERIALS AND METHODS
Eleven larval Ichthyophis kohtaoensis (6.7-14.5 cm total length) were collected near Chiang Dao,
northern Thailand. The animals were anaesthetized in tricaine methanesulfonate (MS 222, Sigma) and unilaterally enucleated. Crystallized horseradish peroxidase (HRP, Boehringer Grade I) was applied to the cut nerve. Care was taken to avoid bleeding and to apply the HRP within 1-2 min following nerve transection. After 2-4 days at room temperature (about 28 °C in Thailand), the animals were deeply anaesthetized and perfused through the ventricle with 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4). The brains were removed, washed several times in 0.1 M phosphate buffer and reacted with the chromogen diaminobenzidine (DAB) as whole- mounts ~6. Seven brains showing label in the optic nerves and tracts were stored in fixative, transferred to our laboratory in Germany, embedded in epoxy
resin, sectioned and viewed with a microscope ~.~sin~ a blue filter (wavelength: 472 nml. In addition, Nissl stained sections of the di- and mesencephahm ol .! Boulengerula houlengeruli. 2 Scoh'comorl~houv kir- kii, and 5 adult lchthyophis kohtaoensis ~crc exam-
ined.
RES U LTS
Description of the di- and mesencephalon This description is based on the results obtained on
the larger specimens of our study. Kuhlenbeck 35,3~ has provided the only available description of the di-
and mesencephalic cell masses of the gymnophionan brain. As in urodeles and lepidosirenid lungfishZ3.4~,, few sulci may indicate different cell masses in the periventricular gray of these brain areas (Fig. 1 ). In accordance with Kuhlenbeck 35, we found 3 majm
sulci along the third ventricle which separates the ep- ithalamus with the almost symmetric habenulae (h) from the dorsal thalamus (dt), the ventral thalamus (vt), and the nucleus preopticus/hypothalamus (np: nomenclature follows Northcutt46). A sulcus limi- tans 35 was not found in our preparations. The separa-
tion of the periventricular gray implied by the sulci was not always accompanied by cytoarchitectural changes. At some places, however, the cells of the
ventral thalamus were smaller than those of the adja- cent dorsal thalamus, or especially the nucleus pre- opticus. In the latter, some larger cells are scattered
inbetween the smaller cells. Gymnophionans posses,, an unusually large subcommissural organ extending from the caudal end of the epiphysis to the posterior commissure (Fig. 6). Further caudally the thala- mus/pretectum (pt) is separated from the tectum me- sencephali by the fiber tracts of the posterior com- missure. No difference in the cytoarchitecture of tha- lamic and tectal neurons appears to separate the thal- amus and tectum. In none of the gymnophionan spe- cies studied (which belong to 3 out of 5 familiesl could we find any indication of the cell layers de- scribed in the tectum by Kuhlenbeck3S. The ventral boundaries of the tectum/tegmentum neurons could not be established with certainty. We assume that the tectal/tegmental junction may extend as far laterally as the most laterally spread retinal efferents. In sum- mary. the periventricular cell masses of the di- and mesencephalon of larval and adult Ichthyophis koh
203
Fig. 1. These line drawings show the visual projection patterns of a larval Ichthyophis kohtaoensis as revealed by anterograde transport of HRP from the right eye. Insets indicate the section levels (A is most rostral) and the position of the eyes. Tracts are indicated by lines, terminal fields by dots, periventricular cell masses by broken lines, and groups of cells by dotted lines. The retinal efferents cross underneath the nucleus preopticus (np) to form superficial and deep fiber tracts. The uncrossed ipsilat- eral fibers run predominantly in a subpial position. Fibers on both sides form a large thalamic neuropil (TN) adjacent to the ventral/dorsal thalamus (vt/dt). Note the few collaterals to the periventricular cells of the dorsal thalamus. Further caudal is a terminal field visible bilaterally underneath the medial optic tract (MeOT). In the pretectum (pt) at the posterior commis- sure (cp) this field (PN) extends laterally underneath the fibers of the marginal optic tract (MaOT). In the tectum opticum (to) only few fibers lateral to the MeOT fibers are found. The basal optic tract (BOT) forms only on the contralateral side a termi- nal field around the entrance of the nervus oculomotorius (BON, n III). Only the contralateral tectal projection extend to the level of the nervus trochlearis (n IV). See list of abbrevia- tions.
taoensis are about as undif ferent ia ted as those of lepi-
dosirenid lungfishes 46 and urodeles 23.
Fiber tracts and terminal fields On enter ing the brain, the ret inal efferents pass
beneath the caudal border of the nucleus preopt icus
(np, Fig. 1). Some collaterals enter the periventr icu-
lar cell masses of the nucleus preopt icus predomi-
nantly on the contra la tera l side. The crossing optic fi-
bers run in superficial and deep fiber bundles within
the ventral white mat ter of the thalamus. The deep ,
per iventr icular fibers may be referred to as the axial
optic tract , whereas the subpial fibers could be called
the tractus opticus marginalis. The ipsilateral fibers
form a loop at the nucleus preopt icus and are almost
exclusively confined to the tractus opticus margina-
lis. On the contra- and ipsilateral sides the fibers pass
through a cluster of cells at the te lencephal ic/dience-
phalic border . Few collaterals enter this cell group.
The marginal and the axial fibers come together on
the contra la tera l side at the level of a thalamic termi-
nal field, the thalamic neuropi l (TN, Figs. 1, 5). This
terminal field is s i tuated lateral to the sulcus medial is
(sm), extending in a dorso-caudal direction. A few fi-
bers form a fiber plexus medial to this terminal field
and reach as far as the per iventr icular cells dorsal to
the suicus medialis (Figs. 1, 5). The thalamic terminal
field is almost equally prominent on the ipsi lateral
side (Figs. 1, 4). However , fewer collaterals to peri-
ventr icular cells were found. No evidence was found
of a segregation of ipsi- and contra la tera l fibers in the
thalamic terminal field.
Dorsal and poster ior to the thalamic terminal field,
the layer of terminals undernea th the fiber tracts be-
comes restr icted to a narrow terminal layer ad jacent
to the more medial fibers of the optic tract. These
medial fibers are t e rmed the medial optic tract
(MeOT) . The more lateral fibers have only few ter-
minals underneath them and are referred to as the
marginal optic tract (MaOT) . Even further lateral ly
a few scat tered fibers form the basal optic t ract
(BOT, Fig. 1). Thus all fibers merge at the thalamic
neuropi l and give rise to 3 tracts: the M e O T (which is
the most prominent) , the MaOT, and the BOT. On
the ipsilateral side the M e O T with its nar row layer of
terminals beneath is clearly present (Fig. 1). The
more lateral fibers of the M a O T , however , are much
sparser than on the contralateral side; only one single
fiber was found in one specimen at the location of an
ipsilateral BOT.
In the dorsal thalamus, alongside the pos ter ior
commissure, an enlargement of the terminal layer
underneath the M e O T forms a pretectal terminal
field, the pretectal neuropi l (PN, Fig. 1). This termi-
nal field extends from the caudal dorsal thalamus,
204
pc
Figs. 2.3 Dorsal views of the brains of an 7.2 cm hmg and a 3.4 cm long lchthyophis lar'~ac In both casc~ t tRP \~a> ,m,'nl~c¢~ " - tht right nervus opticus 72 h prior to sacrifice. The contralateral projection m the prctcctal ncuropil ~ PN t and the tccmm r T,, ,~, ~.¢1 ~, th~ bilateral curse of the retinal efferents in the medial optic tract are revcaled. Note the difference in ipsilater~d pn~lcction~ ~¢, thu tucmn (arrowheads) and the restricted tectal projections. Bar indicates 1 ram.
Figs. 4, 5. The thalamic terminal fields (TN) formed by crossed (Fig. 5) and uncrossed (Fig. 4) retinal cfferent~ are sho~ n in thin curo- hal section. The terminal field is larger on the contralateral side (Fig. 5), receives both subpial and peri~cntricular fibers as ~ell ant shows some collaterals to periventricular cells (arrowhead). Note the label in the ependymal layer (double arrowheadl, Ichttnophi~ larvae, 14.5 cm long, 72 h postoperative survival time.
Fig. 6. The subcommissural organ with the Reisner fibre (arrowhead) of a larval lchthyophis is shown. Compared to other amphibia; orders, gymnophionans possess an unusually large subcommissural organ. Bar indicates 0.1 mm in Figs. 4-(~.
Figs. 7, 8. The pretectum with the ipsilateral (Fig. 7) and the contralateral (Fig. 8) retinal projections are shown in these micrograph~ The pretectal neuropil (PN) underneath the MeOT-fibers is large and bends laterally under the few terminals formed by the MaOT fi. bets. Both tracts and terminal fields are larger on the contralatcral side. Larval lchthyophi~. 14.5 cm Icngth. 72 h postoperative ,ur~ ix. al time, coronal section. Bar indicates 0.1 mm in Figs. 7.8.
205
Figs. 9, 10. These two consecutive cross-sections show the basal optic neuropil (BON) of a larval Ichthyophis (14.5 cm long, 72 h post- operative survival time). Only a very small subpial terminal field surrounds the few fibers. Bar indicates 0.1 mm in Figs. 9, 10.
i.e. the pretectal area, to the rostral tectum. Under-
lying the MeOT fibers, this terminal field extends lat- erally in a sickle shape beneath the MaOT (Figs. 1, 8). On the ipsilateral side, the organization is very similar. The MeOT with the pretectal terminal field is smaller and the MaOT is represented by only few
fibers (Figs. 1, 7). Caudal to the posterior commissure, the pretectal
terminal field becomes less prominent bilaterally be- neath the MeOT fibers. The number of fibers in the MaOT and the MeOT seems to decrease within the tectum in a posterior direction. The MeOT extends
caudally in the tectum, albeit less so on the ipsilateral side (Figs. 2, 3). Compared to the thalamus/pretec- tum, the tectum is almost devoid of terminals from the retina.
At about the level of the entrance of the oculomo- tor nerve, the few fibers forming the BOT on the con- tralateral side turn in a ventral direction to form a narrow, subpial terminal field, the basal optic neuro- pil (BON, Figs. 9, 10). The BON is larger than the tectal terminal field at this level of the brain. No BON was found on the ipsilateral side.
Differences among larvae Only those specimens that showed a dark brown
label on the contralaterai thalamic and pretectal ter- minal fields were used for this study. Three of these specimens were between 6.7 and 7.2 cm long and 4 between 12.5 and 14.5 cm long. All larval Ichthyo- phis may possess an overlapping visual field in the dorsal and rostral direction (Figs. 1, 11). Since the in- testine of the smaller specimens was completely filled with yolk, we conclude that they were posthatching larvae which had not yet started feeding. The larger
specimens were presumably early transformers,
since the smallest fully metamorphosed specimen we collected was 16.5 cm long. In the field, Ichthyophis larvae of about 12-14 cm length stay outside the wa- ter in moist soil during daytime but enter the water for feeding during the night (Himstedt, unpublished observations) thereby resembling the adult in their
life style 7. All larvae examined had well developed bilateral thalamic and pretectal terminal fields and a bilateral MeOT (Figs. 2, 3). The extension of the MeOT within the tectum seemed to be slightly short- er in the caudal direction in the smaller larvae. The MaOT was smaller on the contralateral side and ab- sent on the ipsilateral side of the brains of the smaller larvae. A BOT and a very small BON were found only on the contralateral side.
Artificial labelling Throughout the thalamic projection of retinal fi-
bres, the ependymal cells around the ventricles showed sparse but consistent bilateral labelling with the HRP reaction product (Fig. 5). Since many reti- nal fibers showed signs of degeneration, with fibers being represented only by chains of particles, we as- sume that this ependymal labelling is due to transcel- lular labelling as described in the visual system of anurans TM and teleosts 6.
DISCUSSION
Reliability of the technique Despite earlier reports that HRP cannot migrate in
an anterograde direction, many recent studies sug- gest that this type of labelling is extremely sensitive compared to anterograde degeneration studies and
206
1
11
A
4;
2
4-.
3
B
Fig. 11. The heads with the eyes and the brains with the visual projections (black) from the right eye of adult (A) and larval (B) amphibians are shown in this drawing. In Ichthyophis koh- taoensis both adult (A 0 and larval (Bl) specimens may possess a small dorsal binocular field and show almost identical bilater- al projections of the MeOT. Adult urodeles (A2) and adult anu- rans (A0 show a frontal and a dorsal binocular field, an almost identical contralateral projection to the tectum and a bilateral projection of the MeOT to thalamic and pretectal terminal fields. Posthatching urodeles (B2) but not posthatching anurans (B3) may possess a binocular visual field, which gradually in- creases during larval life in urodeles but develops only around metamorphosis in anurans. The contralateral tectal projection enlarges throughout larval life in both urodeles and anurans (B2,3). However, a small ipsilateral projection develops in uro- deles already in larvae, but enlarges following metamorphosis, Anurans develop an ipsilateral projection only around meta- morphosis. Both binocularity and ipsilateral projections seem to co-vary among anurans perhaps due to thyroxine changes.
even t ransport of radioact ive amino acids 43. In fact,
wholemount prepara t ions using an terograde ly trans-
por ted H R P allows visual projec t ion pa t te rns to be
studied with unsurpassed acuity 16.44. Thus we are
confident that our data reflect quite accurately the
visual project ion pat terns of larval Ichthyophis.
Comparison with adult gymnophionan,~ Our results confirm the finding of a bilateral tha-
lamic project ion in adult lchthyophis 24 and adu[~
Typhlonectes 4. In contrast to the case of adult Typh.. lonectes 4, adult 24 and larval Ichthyophis possess bilat-
eral pretectal and tectal project ions as well as a con-
tralateral project ion of the BON which lies rostral to
the entrance of the oculomotor nerve. In addit ion,
we found a small terminal field at the telencephalic/
thalamic border which may correspond to the nucleus
ovoidales of anurans 67. We also found a small bilater-
al projec t ion to the periventr icular neurons around
the dorsal /ventral thalamus as previously described
in adult urodeles and anurans 15,39 but not in any adult
gymnophionan, plus a very sparse project ion in the
ipsilateral B O T of one specimen. The extent (~f the
larval project ions is even more striking if we consider
that instead of D A B , Himstedt and Manteuffel :a
used the more sensitive chromogen te t ramethyle
benzidine (TMB, ref. 43). This together with the al-
most complete absence of fibers within the M a O T of
adult Ichthyophis 24 suggests that those fibers may be
lost during or following metamorphosis . Loss of fi-
bers has been repor ted in studies dealing with un-
crossed project ions to the super ior colliculus of mam-
malsS,40. Cell death seems to be the major factor un-
derlying this loss in mammals14,28.41
Comparison with urodelian and anuran larvae The most impor tant finding of this comparison is
that in larval Ichthyophis the M e O T with its thalamic
and pretectal terminal fields develops bi lateral ly very
early; compared with the MeOT, the M a O T and its
terminal fields in the tectum are ra ther undifferen-
t iated. In contrast , in both urodeles and anurans the
MaOT develops early, especially to the tectum. The
MaOT increases the tectal terminal field throughout
the larval stage 53.~s, whereas the ipsilateral tract and
terminal areas increase ei ther late in the larval stage
(urodeles, 53) or develop at metamorphos is (anu-
rans, refs. 9, 25, 61). In some amphibians , as with mammals , the deve lopment of the ipsilateral projec-
tion shows some coincidence with binoculari ty, a
state of affairs in contrast to that in reptiles45: (1)
gymnophionans: binocular i ty has not been disproved
and bilateral project ions are present in posthatching
h'hthyophis: (2) urodeles: binoculari ty and bilateral
project ions increase during late larval life. most
prominent around metamorphosis53,59; (3) anurans:
binocularity and bilateral visual projections develop only around metamorphosis (Fig. 11; refs. 9, 25).
We assume this coincidence to represent a co-vari- ance of both binocularity and bilateral projections caused by the same reason, e.g. changes in thyrox- ine25,62. In this context it appears noteworthy that metamorphic changes are only small in Ichthyo- phis 69, and both habitat and life style differ little be- tween larval and adult Ichthyophis 7.
In all amphibians studied to date, the bilateral pro- jections are largely confined to what we have called the MeOT and its related thalamic and pretectal ter- minal fields (presumably homologous to the neuropii of Bellonci and the uncinate field of anurans and uro- deles; refs. 15, 56). The contribution of fibers run- ning in other tracts like the MaOT to the ipsilateral projection seems to be more variable54,55. In anu- rans3S, urodeles (Rettig, in preparation), and mam- malsS, 26, the ipsilateral fibers derive predominantly from the ventro-temporal part of the retina. Whether this is true for gymnophionans as well remains to be examined. In both anurans 56 and urodeles 15 the tha- lamic neuropil of Bellonci shows some segregation between ipsilateral and contralateral fibers; this has been more elegantly demonstrated recently using double labelling techniques 13. No such segregation was found in the presumably homologous thalamic terminal field of Ichthyophis. We assume that this lack of segregation reflects a new acquisition by cae- cilians related to the small number of retinal ganglion cells (Himstedt, in preparation) rather than repre-
sents a primitive character stage 12. The retinal effer- ent-free lateral neuropil of urodeles and anurans15,56 was not observed in Ichthyophis. Since the lateral neuropil is known to be a part of the retino-tectal- thalamic-telencephalic pathway56, its absence may relate to the indistinct tectum. A reduction of retino-
tectal projections and of the retino-tectal-thalamic- telencephalic pathway has been reported in blind snakes 21.
Comparison with other vertebrates A complete absence of contralateral retinal pro-
jections of the tectum has been described to date only in hagfish 37. However, a rather small contralateral projection to the tectum was reported in some lung- fish 46, snakes 21, and in Necturus, a urodele with very
207
small eyes 3. Absence or restriction of tectal terminal fields seems to be a consistent feature of almost all ip- silateral visual projections of vertebrates (selachi- ans3°; polypteriformesS1; ray-finned fisheslO, ll.49.52,60; lungfish47; amphibianslS.24.39.67; turtles2; birds64;
mammalsS,44). Moreover, ipsilateral fibers seem to be confined to the rostro-medial part of the optic tract (MeOT, urodeles and gymnophionans, this study, ref. 15; anterior optic fascicle, birds64; rostral margin of the optic tract, mammals44; medial aspect of the ipsilateral optic tract, reptiles2; tractus opticus dorsomedialis, chondrostei52; fasciculus dorsalis trac- tus opticus, teleosts60). In lampreys, however, the ip- silateral fibres to the tectum are extensive, but not confined to a medial subdivision of the optic tract 33, 34,68. Thus agnathans seemingly lack that bilaterally projecting medial subdivision of the optic tract which appears to be the most strongly developed bilateral tract in all gnathostome vertebrates 2,15.30,47,51,52,e0. Whether this reflects a primitive or a derived condi- tion of agnathan visual projections is not clear at present.
Despite differences in the order of magnitude in numbers of retinal ganglion cells among the 3 am- phibian orders (about 480,000 in anurans42; about 48,000 in urodeleslS; about 4000 in Ichthyophis, Himstedt et al., in preparation) the projections of urodeles and anurans are quite comparablelS.56,67. Similarity of anuran and urodelan visual projections is also implied by recent xenoplastic experiments in which anuran eyes were grafted onto urodeles, and vice versa 22. The almost complete absence of the MaOT as compared to the MeOT in larval and adult Ichthyophis may be due to a selective phylogenetic loss of ganglion cells or collaterals which project in the MaOT. Among amphibians, loss of ganglion cells or collaterals may have led to a comparative enlarge- ment of the bilaterally projecting MeOT, presum- ably by reducing the competition 14 between MaOT and MeOT fibers. A similar trend seems to occur among reptiles, in which the comparatively largest uncrossed thalamic and pretectal projections were reported in the blind snake Typhlops 21. Further stud- ies on tetrapods with a reduced visual system are nec- essary to allow a generalization of this trend at least for tetrapods. This proposal is consistent with the view that nocturnal life style may lead to increased bilateral projections45,57 and the proposed role of the
208
ipsilateral, uncrossed projections in intensity dis-
crimination 5,21,24. In contrast to tetrapods, data on
lungfish 4" and hagfish 37 indicate that numerical re-
duction of retinal ganglion cells is not necessarily cor-
related with a relative enlargement of the uncrossed
retinal projection, but may lead to a complete loss of
the ipsilaterally projecting ganglion cells or collate-
rals as compared to other species of these groups3a.47.
In summary, lchthyophis kohtaoensis larvae are
unique among amphibians in possessing a very early
bilateral retinal projection which shows little in-
crease during the larval stage. They may undergo a
selective loss of retinal axons in the MaOT around or
following metamorphosis. As in other amphibians,
there may be some coincidence between binocularity
and bilateral projections in the larvae, but hin,~culzt~.
ity of the dorsal visual field remains to be pr~wcn.
Whether the ipsilateral projection derives ~.'xclusivc-
ly from the temporal retina as in other amphibians
and mammals 2(' remains to be studied.
ACKNOWLEDGEMENT
We wish to express our sincere thanks to Dr. D.
Forsythe for her help with the English and to Drs.
Marvalee H. Wake, G. Manteuffel, G. Rettig, and
Mr. H. Wicht for helpful suggestions on an earlier
draft of this manuscript. This work was supported by
n III nucleus oculomotorius n IV nucleus trochlearis pc posterior commissure PN pretectal neuropil pt pretectum sh sulcus hypothalamicus sm sulcus medialis tel telencephalon tm tegmentum TN thalamic neuropil to tectum opticum vt ventral thalamus
REFERENCES
1 Ariens-Kappers, C.U., Huber, G.C. and Crosby, E.C., The Comparative Anatomy of the Nervous System of Verte- brates, including man, Hafner, New York, 1936.
2 Bass, A.H. and Northcutt, R.G., Retinal recipient nuclei in the painted turtle, Chrysemys picta: an autoradiographic and HRP study, J. Comp. Neurol., 199 (1981) 97-112.
3 Caldwell, J.H. and Berman, N., The central projections of the retina in Necturus maculosus, J. Comp. Neurol., 171 (1977) 455-464.
4 Clairambault, P., Cordier-Picouet, M.-J. and Pairault, C., Premieres donn6es sur les projections visuelles d'un Am- phibien Apode (Typhlonectes compressicauda), C.R. Acad. Sci. Set. D., 291 (1980) 283-286.
5 Cowey, A. and Franzini, C., The retinal origin of uncrossed optic nerve fibres in rats and their role in visual discrimina- tion, Exp. Brain. Res., 35 (1979) 443-455.
6 Crapon de Caprona, M.-D. and Fritzsch, B., The devel- opment of the retinopetal nucleus olfacto-retinalis of two cichlid fish as revealed by horseradish peroxidase, Dev. Brain Res., 11 (1983) 281-301.
7 Crapon de Caprona, M.-D. and Himstedt, W., Das aqua- tische Verhalten der Blindw~ihle Ichthyophis kohtaoensis, Taylor 1960 (Amphibia: Gymnophiona), Salamandra. in press.
8 Cusick, C.G. and Kaas, J.H., Retinal projections in adult and newborn grey squirrels, Dev. Brain Res., 4 (1982) 275-284.
9 Currie, J. and Cowan, W.M., Evidence for the late devel- opment of the uncrossed retinothalamic projections in the frog, Ranapipiens, Brain Res., 71 (1974) 133-139.
10 Dunn-Meynell, A.A., Prasada Rao, P.D. and Sharma, S.C., The ipsilateral retinotectal projection in normal and albino channel catfish, Neurosci. Lett., 36 (1983) 25-31.
l l Ebbesson, S.O.E. and Ito, H., Bilateral retinal projections in the black piranah (Serrasalmus niger), Cell Tissue Res.. 213 (1980) 483-495.
12 Ebbesson, S.O.E., Evolution and ontogeny of neural cir- cuits, Behav. Brain Sci., 7 (1984) 321-366.
13 Ebbesson, S.O.E. and Bazer, G.T., Double labeling of neural circuits using horseradish peroxidase and cobalt, Neurosci. Lett., submitted for publication.
14 Fawcett, J.W., O'Leary, D.D. and Cowan, W.M., Activity and the control of ganglion cell death in the rat retina, Proc. Natl. Acad. Sci. U.S.A., 81 (1984)5589-5593.
.15 Fritzsch, B., Retinal projections in European Salamandri- dae, Cell Tissue Res., 213 (1980) 325-341.
16 Fritzsch, B. and Nikundiwe, A.M., Tracing neuronal con- nections in whole mounted brains of small vertebrates. Mi- croscopy, 41 (1984) 145-149.
17 Frost, D.O., Axonal growth and target selection during de-
velopment: retinal projections to the ventrobasal complex and other 'nonvisual' structures in neonatal syrian hamster, J. comp. Neurol., 230 (1984) 576-592.
18 Gruberg, E.R., Optic fiber projections of the tiger sala- mander Ambystoma tigrinurn, J. Hirnforsch., 14 (1972) 399-411.
19 Guillery, R.W., Sitthi Amorn, C. and Eighmy, B.B., Mu- tants with abnormal visual pathways: an explanation of anomalous geniculate laminae, Science, 174 (1971) 831-832.
20 Guillery, R.W. and Updyke, B.V., Retinofugal pathways in normal and albino axolotls, Brain Res., 109 (1976) 235-244.
21 Halpern, M., Retinal projections in blind snakes, Science, 182 (1973) 390-391.
22 Harris, W.A. and Cole, J., Common mechanisms in verte- brate axonal navigation: retinal transplants between dis- tantly related amphibians, J. Neurogenetics, 1 (1984) 127-140.
23 Herrick, C.J., The Brain of the Tiger Salamander Ambysto- rna tigrinum, University of Chicago Press, Chicago, 1948.
24 Himstedt, W. and Manteuffel, G., Retinal projections in the caecilian lchthyophis kohtaoensis (Amphibia, Gymno- phiona), Cell Tissue Res., 239 (1985) 689-692.
25 Hoskins, S.G. and Grobstein, P., Induction of the ipsilater- al retinothalamic projection in Xenopus laevis by thyrox- ine, Nature (London), 307 (1984) 730-733.
26 Hsiao, K., Bilateral branching contributes minimally to the enhanced ipsilateral projection in monocular syrian ham- sters, J. Neurosci., 4 (1984) 368-373.
27 Humphrey, R.R., Albino axolotls from an albino tiger sala- mander through hybridization, J. Hered., 58 (1967) 95-101.
28 Insausti, R., Blakemore, C. and Cowan, W.M., Ganglion cell death during development of ipsilateral retino-collic- ular projection in golden hamster, Nature (London), 308 (1984) 362-365.
29 Jen, L.S. and Lund, R.D., Experimentally induced en- largement of the uncrossed retinotectal pathway in rats, Brain Res., 211 (1981) 37-57.
30 Jen, L.S., So, K.-F., Yew, D.T. and Lee, M., An autoradi- ographic study of the retinofugal projections in the shark, Hemiscyllium plagiosurn, Brain Res., 274 (1983) 135 - 139.
31 Jen, L.S., So, K.-F. and Chang, A.-B., An anterograde HRP study of the retinotectal pathways in albino and pig- mented guinea pigs, Brain Res., 263 (1983) 331-335.
32 Jen, L.S., So, K.-F. and Woo, H.H., An anterograde HRP study of the retinocollicular pathways in normal hamsters and hamsters with one eye enucleated at birth, Brain Res., 294 (1984) 169-173.
33 Kennedy, M.C. and Rubinson, K., Retinal projections in larval, transforming and adult sea lamprey, Petromyzon marinus, J. comp. Neurol., 171 (1977) 465-480.
34 Kosareva, A.A., Retinal projections in lamprey (Lampetra fluviatilis), J. Hirnforsch, 21 (1980) 243-256.
35 Kuhlenbeck, H., Zur Morphologie des Gymnophionenge- hirns, Jena. Z. Naturwiss., 58 (1922) 453-484.
36 Kuhlenbeck, H., Malewitz, T.D. and Beasley, A.B., Fur- ther observations on the morphology of the forebrain in gymnophiona, with reference to the topologic vertebrate forebrain pattern. In R. Hassler and H. Stephan (Eds.), Evolution of the Forebrain, Plenum, New York, 1967, pp. 9-19.
37 Kusunoki, T. and Amemiya, F., Retinal projections in the
38 Lazar, G., The projection of the retinal quadrants on the optic centres in the frog, Acta Morphol. Acad. Sci. Hung., 19 (1971) 325-334.
39 Levin, R.L., An autoradiographic study of the retinal pro- jection in Xenopus laevis with comparisons to Rana, J. comp. Neurol., 189 (1980) 1-29.
40 Manford, M., Campbell, G. and Lieberman, A.R., Postna- tal development of ipsilateral retino-geniculate projections in normal albino rats and the effects of removal of one eye at birth, Anat. Embryol., 170 (1984) 71-78.
41 Martin, P.R., Sefton, A.J. and Dreher, B., The retinal lo- cation and fate of ganglion cells which project to the ipsilat- eral superior colliculus in neonatal albino and hooded rats, Neurosci. Lett., 41 (1983) 219-226.
42 Maturana, H.R., Number of fibers in the optic nerve and ~ the number of ganglion cells in the retina of anurans, Nature (London), 183 (1959) 1406-1407.
44 Murabe, Y., Fujisawa, H., Terubayashi, H. and Ibata, Y., A whole-mount horseradish peroxidase study of the retinal central projection in normal and monocular rats, Anat. Em- bryol., 168 (1983) 151-158.
45 Northcutt, R.G. and Butler, A.B., Evolution of reptilian visual systems: retinal projections in a nocturnal lizard, Gekko gecko (Linnaeus), J. Comp. Neurol., 157 (1974) 453-466.
46 Northcutt, R.G., Retinofugal projections in the lepidosire- nid lungfishes, J. Cornp. Neurol., 174 (1977) 533-574.
47 Northcutt, R.G., Retinal projections in the Australian lungfish, Brain Res., 185 (1980) 85-90.
48 Northcutt, R.G. and Wathey, J.G., Guitarfish possess ipsi- lateral as well as contralateral retinofugal projections, Neu- rosci. Lett., 20 (1980) 237-242.
49 Northcutt, R.G., Evolution of the optic tectum in ray- finned fishes. In R.E. Davis and R.G. Northcutt (Eds.), Fish Neurobiology, University of Michigan Press, Ann Ar- bor, 1983, pp. 1-42.
50 O'Leary, D.D.M., Gerfen, C.R. and Cowan, W.M., The development and restriction of the ipsilateral retinofugal projection in the chick, Dev. Brain Res., 10 (1983) 93-109.
51 Reperant, J., Rio, J.P., Miceli, D., Amouzou, M. and Pey- richoux, J., The retinofugal pathways in the primitive Afri- can bony fish Polypterus senegalus (Cuvier, 1829), Brain Res., 217 (1981) 225-243.
52 Reperant, J., Vesselkin, N.P., Ermakova, T.V., Rusta- mov, E.K., Rio, J.P., Palatnikov, G.K., Peyrichoux, J. and Kasimov, R.V., The retinofugal pathways in a primitive ac- tinopterygian, the chondrostean Acipenser gueldenstaedti. An experimental study using degeneration, radioautogra- phic and HRP methods, Brain Res., 251 (1982) 1-23.
53 Rettig, G., Fritzsch, B. and Himstedt, W., Development of retinofugal neuropil areas in the brain of the alpine newt, Triturus alpestris, Anat. Embryol., 162 (1981) 163-171.
54 Rettig, G. and Roth, G., Afferent visual projections in three species of lungless salamanders (Family Plethodonti- dae), Neurosci. Lett., 31 (1982) 221-224.
55 Rettig, G. and Roth, G., Retinofugal projections in sala- manders of the family Plethodontidae, Cell Tissue Res., submitted for publication.
56 Scalia, F., The optic pathway of the frog: nuclear organiza- tion and connections. In R. Llinas and W. Precht (Eds.),
210
Frog Neurobiology, Springer, Berlin, 1976, pp. 386-406. 57 Schroeder, D.M., Retinal afferents of an infrared sensitive
snake, Crotalus viridis, J. Morphol., 170 (1981) 29-42. 58 Scott, T,M., The development of the retino-tectal projec-
tion in Xenopus laevis: an autoradiographic and degenera- tion study, J. Embryol. Exp. Morphol., 31 (1974)409-414.
59 Sivak, J.G. and Warburg, M.R., Optical metamorphosis of the eye of Salamandra salamandra, Can. J. Zool., 58 (1980) 2059-2064.
60 Springer, A.D. and Gaffney, J.S., Retinal projections in the goldfish: a study using cobaltous-lysine, J. Cornp. Neu- rol., 203 (1981) 401-424.
61 Steedman, J.G., Stifling, R.V. and Gaze, R.M., The cen- tral pathways of optic fibres in Xenopus tadpoles, J. Em- bryol. Exp. Morph., 50 (1979) 199-215.
62 Stirling, R.V. and Br/indle, K., Expansion of the visual pro- jection to the tectum of axolotls during metamorphosis, Dev. Brain Res., 5 (1982) 343-345.
63 Straznicky, C. and Hiscock, J., Aberrant retinotectal pro- jection induced by larval unilateral enucleation in Xenopus, Neurosci. Lett., 39 (1983)5-10.
64 Takatsuji, K., Ito, H. and Masai, H., Ipsilateral retinal pro- jections in Japanese quail, Coturnix coturnix japonica,
Brain R es. Bull., 10 (1983) 53-56. 65 Thanos, S. and Bonhoeffer, F., Development ~{ the tran.
sient ipsilateral retinotectal projection in the chick embryo: a numerical fluorescence-microscopic analysis..I (~,mp Neurol., 224 (1984) 407-414.
66 Thanos, S., Fujisawa, It. and Bonhoeffer. F . Elimination of ipsilateral retinotectal projections in mono-ophthalmic chick embryos, Neurosci. Lett., 44 (1984) 143-14g.
67 Toth, P., Lazar, G. and Goercs, T., Retinal projections m the adult Xenopus laevis: a study with cobalt filling tech- nique, Acta Morph. Acad. Sci. Hung., 28 (198(t) 365- 374.
68 Vesselkin, N.P,, Ermakova, T.V., Reperant, .1., Kosareva, A.A. and Kenigfest, N.B., The retinofugal and retinopetal systems in Lampetra fluviatilis. An experimental study using radioautographic and HRP methods. Brain Re~, 195 (1980) 453-460.
69 Welsch, U., Schubert, C. and Storch, V., Investigations on the thyroid gland of embryonic, larval and adult lchthyo- phis glutinosus and lchthyophis kohtaoensiv. ('ell Tivs'ue Res., 155 (1974) 245-268.
70 Wilczynski, W. and Zakon, H., Transcellular translcr ol HRP in the amphibian visual system. Brain Res., 239 (1982) 29-40.
