Identification of receptors on the lip and central neurons innervating peripheral chemoreceptive structures of Lymaea stagnalis

Material Information

Identification of receptors on the lip and central neurons innervating peripheral chemoreceptive structures of Lymaea stagnalis
Nelson, Gina Marie
Publication Date:
Physical Description:
124 leaves : illustrations ; 28 cm


Subjects / Keywords:
Gastropoda ( lcsh )
Neural circuitry ( lcsh )
Gastropoda ( fast )
Neural circuitry ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 101-106).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Integrative Biology.
Statement of Responsibility:
by Gina Marie Nelson.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
19719611 ( OCLC )
LD1190.L45 1985m .N44 ( lcc )

Full Text
Gina Marie Nelson
B.A., University of Colorado, 1982
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Department of Biology

This thesis for the Master of Arts degree by
Gina Marie Nelson
has been approved for the
Department of
/L)s 0, (> / f / J

Nelson, Gina Marie (M.A., Biology)
Identification of Receptors on the Lip and Central
Neurons Innervating Peripheral Chemoreceptive
Structures of Lymnaea Stagnalis
Thesis directed by Assistant Professor Teresa E.
The purpose of this study is to identify
neural elements of the snail Lymnaea stagnalis
which may be involved in chemoreception. The study
is a two-part investigation of the neurons in the
central nervous system (CNS) which have processes in
nerves that innervate sensory structures and the
receptors on the surface of one of these structures.
The first part of the study is mapping the
neurons in the CNS which send axons through nerves
innervating the lips and tentacles. The neurons,
identified by using a cobalt backfilling technique,
may participate in the neural circuitry involved in
receiving and processing chemical and mechanical
stimuli, and producing motor outputs to the lip and
tentacle muscles. At least some of the cells are
those involved in the chemoreceptive circuitry. The
method, though, does not differentiate between
different functions. The cells are indicated on a
series of maps.

The second part of the study identifies
receptors on the lip of Lymnaea, one of the sensory
structures. These receptors are identified by five
criteria: structure, morphological appearance of the
nucleus, location, presence of cilia, and the
receptor candidates' similarity to receptors
identified in other studies. These receptors were
investigated by utilizing light and electron
microscopy. New, previously unidentified receptors
are identified, and previously identified receptors
are confirmed in this study.
The two parts of the study are tied together
by a model of the wiring between the neurons in the
CNS and the receptors. Four possible patterns of
wiring are presented. In order to determine these
patterns, a cobalt backfilling technique, similar to
that used in part one, was utilized in this part.
It is intended that the information here be
utilized by other investigators to discern more
precisely by physiological methods, the chemoreceptive
circuitry of Lymnaea stagnalis.

I. GENERAL INTRODUCTION........................... 1
Summary of Study............................. 3
II. INTRODUCTION TO PART ONE...................... 8
III. METHODS AND MATERIALS......................... 11
IV. RESULTS....................................... 15
Nomenclature................................ 15
General Features............................ 34
Special Cases............................... 36
V. DISCUSSION.................................... 39
Variability............................. 3 9
Other Studies............................... 41
Summary..................................... 43
VI. INTRODUCTION TO PART TWO..................... 44
Vertebrate vs. Invertebrate Sensory
Epithelium............................... 44
VII. METHODS AND MATERIALS......................... 46
VIII. RESULTS....................................... 49
Overall Morphology of Head Skin............ 49
Distribution of Epithelial Regions of
the Lip............................... . 6 2
Criteria for Identification of
Receptors................................ 63
Previously Unidentified Receptors.... 67

VIII. Continued
Receptors Previously Identified.......... 72
Basal Cell............................... 75
Unspecialized Epithelium.................... 76
Cobalt Backfilling......................... 77
IX. DISCUSSION.................................... 82
Previously Unidentified Receptors........... 84
Comparison to Other Studies................. 85
Basal Cells................................. 88
Curly Microvilli............................ 90
Distribution of Cells....................... 91
Function of Receptors....................... 92
Vertebrate and Invertebrate................. 95
Form and Function of Chemoreceptors... 95
Cobalt Backfilling: An Overall Model.... 98
BIBLIOGRAPHY...................................... 101
A. CELL DATA.................................... 108

1. Cells From Different Maps That May Be the
Same But Could Not Be Positively
Identified..................................... 35
2. Cell Data..................................... 108
3. Comparison of Cells in This and Other
Studies........................................ 42
4. Cells in Lymnaea Lip Epithelium................ 69

1. Theory on Arrangement of Neurons............. 9
2. Plastic Dish for Backfilling Neurons......... 12
3. Cerebral-Buccal Connective Fill Maps
Part a....................................... 18
Part b....................................... 19
Part c....................................... 20
Part d....................................... 21
4. Tentacle Nerve Fill Maps
Part a....................................... 22
Part b....................................... 23
Part c....................................... 24
Part d....................................... 25
5. Superior Lip Nerve Fill Maps
Part a....................................... 26
Part b....................................... 27
Part c....................................... 28
Part d....................................... 29
6. Median Lip Nerve Fill Maps
Part a...................................... 30
Part b....................................... 31
Part c....................................... 32
Part d

7. Dorsal and Ventral Views of Lymnaea
Head.......................,.............. 50
8. Epithelial Regions............................ 51
9. Ganglion-rich Region of Lip Epithelium........ 52
10. Ganglion-rich Region of Lip Epithelium........ 53
11. Ganglion-rich Region Near Groove.............. 54
12. Ganglion Processes in Lip Epithelium.......... 55
13. Receptor Ending E6............................ 56
14. Plain Region of Lip Epithelium................ 58
15. Plain Region of Lip Epithelium................ 59
16. Mucous Region of Lip Epithelium............... 60
17. Unspecialized Epithelium, From Head Skin.... 61
18. Characteristic Drawings of Receptors.......... 68
19. Receptor Ending 2............................. 71
20. Cobalt Backfill Control....................... 78
21. Cobalt Backfilled GRR Epithelium.............. 79
22. Light Microscope Photograph of Cobalt Filled
Epithelium.................................. 80
23. Model of Lymnaea Wiring....................... 99

In order to investigate mechanisms in plasti-
city the circuitry of the neurons involved must be
understood. Plasticity is a change in the nervous
system due to experience. Classical conditioning
leads to an alteration in the behavior of an indi-
vidual, being mediated by plasticity. Learning,
however, can only occur if the nervous system stores
past information from the environment, allowing the
animal to recall its past experience and modify its
future behavior because of it. Alexander and
Audesirk (Alexander, Audesirk, and Audesirk, 1982,
1984; Audesirk, Alexander, and Audesirk, 1982)
showed that Lymnaea are capable of displaying a
modification in their behavior, i.e.: they are
able to learn. Experience teaches them to react
when presented with a stimulus (this stimulus is
termed a conditioning stimulus), when in the past
the same stimulus was associated with food.
The action that occurs in this situation is
the reception of a chemical stimulus, processing in
the central nervous system, and a motor output.

These actions are mediated by chemoreceptors, neurons
in the central nervous system, and motor neurons. The
organization of these components needs to be under-
stood if learning is to be described physiologically.
The most important part of this system to be studied
in terms of learning is the chemoreceptor and the
central nervous system (CNS) neurons that they are
connected to. These two elements combined can be
referred to as the chemoreceptive circuitry.
A model of the wiring connections of Lymnaea
then needs to be developed which points out the
components in this chemoreceptive circuitry, the
peripheral receptors, the neurons in the CNS, and the
connections between them. Once this is accomplished
the changes that take place when learning occurs in
Lymnaea can begin to be investigated. Thus, the
purpose of this study is to provide a model of
Lymnaea chemoreceptive circuitry, identifying the
elements of the circuitry.
The elements of the chemoreceptive circuitry
can be studied in two parts: first, the central
nervous system and second, the sensory receptors in
contact with the environment. The first requires
identifying the neurons in the CNS which are involved
in processing particular types of information and
outputting responses. (A unique characteristic of

gastropods is that individual cells can be identified
by location, electrical, and physical characteristics.
This is not possible in mammals.) The second involves
identifying in detail the receptors and circuitry
involved in transmitting information between the
environment and the central nervous system. The
overall purpose of this study is to provide one large
model involving both of these parts of Lymnaea1s
chemoreceptive circuitry.
Summary of Study
The first part of this study involved mapping
the neurons in the central nervous system which send
axons through nerves innervating the lips and
tentacles. Bovjerg (1968) described these are the
major food-responsive organs. They can also be
described as "sensory structures." The neurons,
identified by using a cobalt backfilling technique,
may participate in the neural circuitry involved in
receiving and processing chemical and mechanical
stimuli, and motor output to lip and tentacle muscles.
At least some of the cells are those involved in the
chemoreceptive circuitry.
Benjamin and associates (Benjamin and Rose,
1979, 1980; McCrohan and Benjamin, 1980a, 1980b; Rose
and Benjamin, 1981a, 1981b) have described in

detail neurons involved in the motor aspects of
feeding in Lymnaea. Other mapping studies include
Lever's work (1977) on identified neurons involved in
tentacle contraction; Winlow and Benjamin (1976)
mapped cells in the parietal, visceral, and pedal
ganglia; Slade et al (1981) mapped cells of the pedal
ganglia. Mapping studies on other gastropods include
those of Helix aspersa (Kerkut et al, 1975, and Par-
mentier, 1973), Aplysia (Fredman and Jahan-Parwar,
1975), Planorbis corneus (Sonetti et al, 1982),
Haementeria ghiliani (Kramer and Goldman, 1981), and
Archachatina marginata (Nisbet, 1960).
Together these studies identify small collec-
tions of neurons from the CNS that are involved in
mediating some behavior. For instance, Lever's work
(1977) outlined the CNS neurons that are involved in
tentacle contraction when the tentacle is stimulated.
These cells which are identified by Lever are a small
part of the cells involved in a large system which,
up until this point, has never been presented as a
whole. The other studies also identify small compo-
nents of sensory circuitry, but they are either
limited to specific behaviors or to specific loca-
tions in the CNS.
What this study does is to provide a whole
array of maps that encompass the major chemoreceptive

nerves and almost all of the neurons which have
processes in them. Some cells, with a particular
structural arrangement, are not stained. In
particular, interneurons and primary chemoreceptors
do not stain. Thus, in theory, nearly every cell,
both chemosensory and mechanosensory, which is
involved in plasticity when the snail experiences
chemosensory stimulation, is identified. With the
exception of the above mentioned cell types, it is a
relatively complete localization of any cells that
may be involved in the circuitry of chemoreception
and reaction.
The second part of this study identifies
some of the receptor endings that are connected to
the neurons in Part I. These receptors are in
contact with the environment and are located on the
lip of Lymnaea, one of the food-responsive organs.
(The lip is defined as the area surrounding the mouth
of the ventral side of the snail's head, or the
ventral side of the oral lappet. It is most easily
seen when the snail is extended in a crawling or
feeding position.)
Zylstra (1972a) describes sensory endings in
several locations on Lymnaea1s head, categorizing
them by structures extending from the apical surface
(i.e., cilia and microvilli), and by the structure of

ciliary roots. Zaitseva (1981) categorized endings on
the head of Lymnaea based on scanning electron micro-
graphs and Golgi staining results. Both of these
studies limited descriptions to apical structures of
the receptor cells. Other studies describing gastro-
pod sensory receptors include work on Nassarius
reticulatus (Crisp, 1971), the lip region (Benedeczky,
1971) and tentacles (Wondrak, 1981) of Helix pomitia
L., Aplysia (Emery and Audesirk, 1978), Pleurobranchaea
californica (Davis and Matera, 1982), and the slug
Arion ater (Wright, 1974).
This study describes the entire lip as a
sensory organ. It identifies areas of the epithelium
which are most likely involved in stimuli reception
and areas which are less likely involved. It also
describes cell types which may function as receptors
and important nonreceptor cells associated by loca-
tion, which may have a role in sensory reception and
information processing.
While considering the elements that are part
of the chemosensory circuitry from the level of the
CNS and the level of the peripheral elements it is
important not to leave out how they are connected
together. This information is needed to form a
complete model of the chemosensory circuitry. A
preliminary model is presented which takes into account

all the possible connections between the two levels
and the structural possibilities that may occur at
each level. Part of this study was to eliminate some
of these possibilities and to suggest the wiring
patterns from actual observations. The results from
studying these connections tie the other two sections
of the study together.

Maps of the neurons in the CNS which have
connections to chemoreceptive organs provide a place
to start for detailing the circuitry involved in
chemoreception. Backfilling chemoreceptive organ
nerves with cobalt reveals the neurons which have
associations to the periphery as chemoreceptors,
raechanoreceptors, interneurons, or motorneurons.
Figure 1 illustrates all the theoretically
possibly connections (based on previous studies by
other investigators) between cells, the types of
receptors that may be in the periphery, and how they
may be connected to the CNS neurons. (Each possibility
is shown at each level but not every possible combi-
nation is drawn.) At the peripheral level there are
three possibilities. A: cell bodies (open circles)
may be located under the basal lamina and send
processes to form primary receptor endings in the
epithelial layer. (Al: cell bodies randomly arranged
in the subepithelium. A2: cell bodies in a ganglia,
a collection of cell bodies.) B: receptors may occur
singly, without any distinct association to other

Figure 1. Theory on Arrangement of Neurons.
The theory as to how neurons may be arranged in the
three areas of the snail's body (epithelium, subepi-
thelium, and central nervous system (CNS)) and how
they may be connected together. The dark and open
circles represent the action of the cobalt. See text
for explanation.
neurons. In other words, they may occur randomly
throughout the epithelium, not gathered in a partic-
ular group. C: the primary receptors may be totally
intraepithelial, either sending down a process to the
central level or having secondary synapses on it at
the level of the epithelium.
At the central level there are basically the
following three possibilities. D: cells in the
central nervous system may send processes out the
nerves innervating the receptor organs. (Dl:

synapsing at the base of the epithelium. D2:
synapsing in a ganglion.) E: primary receptors
may synapse on neurons in the CNS. F: processes
extending from the cells at both levels may synapse
at some location in between. G: one last organiza-
tional aspect to consider is that there may be inter-
mediate cells between the epithelial primary receptors
and the cells in the CNS. The dark circles represent
cells which would fill with cobalt. The open ones,
because of the location of their synapses/ would not
be filled with cobalt, and therefore would not show
up on the maps.

Lymnaea stagnalis, derived from stock orig-
inating in England, were raised in the lab in fish
tanks and fed Tetra-min SM-80 fish food. Specimens
ranging from 1.23 to 2.5g and averaging 1.8g were
selected for study. Larger snails had too much
connective tissue around the ganglion, a condition
which made identifying nerves difficult. Smaller
snails were difficult to work with because of their
size. The brains were dissected out in a Lymnaea
Ringers solution (NaCl, 55mM; CaC^/ 4mM; KCl, 2mM;
MgCL2, 2mM; tris buffer, 4mM; glucose, 10mg/l; pH 7.8,
modified from Benjamin, 1978 and the data of Soffe,
Benjamin, and Slade, 1980).
Cobalt backfilling allows introduction of a
substance into the neurons via the processes they
extend into the nerve. With time the cobalt diffuses
from the capillary, down the process in the nerve,
and into the cell body. To accomplish this cut ends
of nerves were filled using a plastic dish which had
capillary tubes, with their ends drawn to tips of
various sizes, inserted through its walls (Figure 2).

Figure 2. Plastic Dish for Backfilling Nerves.
For backfilling nerves, a plastic Petri dish was lined
with clear encapsulating resin, and holes were punched
in the side with a hot needle. Capillary tubes with
heat-drawn tips of various sizes were inserted through
the holes and anchored to the dish with silicone
rubber glue.
This design was first developed by Dr. B. Granzow.
The excised brain was placed in the dish and covered
with Ringers. Care was taken not to touch or squeeze
the selected nerve so that the flow of dye into the
axons would not be obstructed. The brain was then
moved to a capillary of the appropriate size, and the
nerve was drawn in by suction from a syringe attached

to the other end of the capillary by plastic tubing.
Ringers was then removed from the capillary with a
syringe needle to insure that the tip was sealed.
Cobalt hexamine chloride (Co(NH^)gCl^ r .025M) was
injected into the capillary with the syringe needle
until the entire capillary was filled. The dye was
then left to diffuse into the nerve for 18-24 hours
at 4C. With time the cobalt diffuses from the
capillary, down the process in the nerve, and into
the cell body.
Brains were removed from the backfill chamber
pinned out on silicon resin in a Petri dish and the
pedal commissure was cut. The cobalt was precipitated
with ammonium sulfide (7% v/v) for 8-10 minutes. The
brains were then fixed for 25 minutes in 4% gluter-
aldehyde. It was noted that the longer the tissues
were exposed to gluteraldehyde, the darker they became.
Fixing them overnight made them so dark that locating
stained cell bodies became difficult. However,
eliminating the exposure to gluteraldehyde rendered
the ganglia quite clear so that identifying the bound-
aries of the ganglia became difficult. A fixation
time of 25-30 minutes fixed the tissues and allowed
enough color to facilitate mapping.
Tissues were then dehydrated in ethylene glycol
monomethylether (EGME), followed by 100% propanol, and

cleared in two changes of glycol methacralate (GM).
The brain was placed in GM on a slide and covered with
a coverslip. Exposure to UV light for approximately
15 minutes or so polymerized the GM. Wholemount slides
were projected onto drawing paper and the outline was
drawn. Stained cell bodies were mapped using a dis-
secting microscope. Individual maps were made of the
dorsal and ventral faces of each of 20 backfills, (10
filled from the right, 10 from the left) for each of
the following nerves: superior lip nerve (SLN),
median lip nerve (MLN), tentacle nerve (TN), and
cerebrobuccal connective (CBC). (These nerves inner-
vate the chemoreceptive organs, as mentioned before.
The TNs innervate the tentacles. The MLNs and SLNs
innervate the lip region. The CBC is the connective
between the cerebral and buccal ganglia. The latter is
comprised of neurons primarily involved in controlling
the movement of the muscular buccal mass, which moves
the mouth.) Composite maps for each nerve were then
drawn by comparing all the individual maps for each
nerve and each side.

A system of nomenclature was devised in which
each capital letter or number in the name represents
some feature that could be used in identifying a cell.
This combination of letters and numbers results in a
unique name for each cell or group of cells and avoids
confusion. Aspects of this system are similar to
those employed by Sonetti (1982) and Parmenter (1973).
Group and individual names consist of a series of six
abbreviations. The first letter indicates the nerve
that was filled: S, superior lip nerve; M, median lip
nerve; L, both superior and median lip nerve; T,
tentacle nerve; C, cerebrobuccal connective. (In some
cases it appeared that the same neurons were filled
from both the MLN and SLN.) Some of these cases were
confirmed using intracellular staining with Lucifer
Yellow, a fluorescent dye (W. W. Stewart, 1978). The
second letter indicates the face of the brain on
which the cells are located: D, dorsal; V, ventral.
The third letter indicates the side of the brain on

which the cell is located: R, right; L, left. Since
there is only one visceral ganglion the R and L prefix
is omitted. The fourth position indicates the ganglion
in which the cell is located: C, cerebral; B, buccal;
Pi, pleural; P, parietal; Pe, pedal; V, visceral. The
fifth place is held by a number for individual cells
or a letter for clusters (a capital and lower case
letter designate different groups). After the dash
are two informational abbreviations. The first abbre-
viation tells the size of the cell or average size of
cells in the cluster: S, small (<30um); M, medium
(30-70um); L, large (>70um). The last designation
applies only to individual cells and indicates whether
they are on the surface of the ganglion (srf) or are
sub-surface cells (sub). This does not apply to groups
because members of the same group could be spread from
the surface to the sub-surface. On the maps, surface
cells are represented by solid and subsurface cells by
hollow circles.
Dorsal and ventral maps were generated for
the left and right members of each of the paired
nerves (Figures 3-6). When homologous pairs of neurons
send axons out both left and right nerves they are
given the same name. For example, the cell MDLCll-M,
srf is located in the left cerebral and MDRCll-M,srf is
located in the right cerebral, both on the dorsal face

CDLCm-Mn 5
2 4
Figure 3. Cerebral-Buccal Connective Fill Maps,
a) CBC-filled from the left side, dorsal view.
Locator letters are the first four letters of the
label (three in the case of the visceral ganglia).
x: Also known as Pi.

Q-S 3-M 0-S
Figure 3 (continued). Cerebral-Buccal Connective
Fill Maps.
b) CBC-filled from the right side, dorsal view.
Locator letters: left buccal, CDLB; right buccal,
CDRB; left cerebral, CDLC; right cerebral, CDRC;
left pedal, CVLPe; right pedal, CVRPe; left pleural,
CDLPl; right pleural, CDRPl; left parietal, CDLP;
right parietal, CDRP; visceral, CDV; x, also known
as AlO.

Figure 3 (continued). Cerebral-Buccal Connective
Fill Maps.
c) CBC-filled from the left side, ventral view.
(CG): cerebral giant.

Figure 3 (continued). Cerebral-Buccal Connective
Fill Maps.
d) CBC-filled from the right side, ventral view.
Locator letters: left buccal, CVLB; right buccal,
CVRB; left cerebral, CVLC; right cerebral, CVRC;
left pedal, CDLPe; right pedal, CDRPe; left pleural,
CVLPl; right pleural, CVRPl; left parietal, CVLP;
right parietal, CVRP; visceral, CW; (CG) cerebral

Figure 4. Tentacle Nerve Fill Maps,
a) TN-filled from the left side, dorsal view.
Locator letters: left buccal, TDLB; right buccal,
TDRB; left cerebral, TDLC; right cerebral, TDRC;
left pedal, TVLPe; right pedal, TVRPl; left
pleural, TDLPl; right pleural, TDRPl; left parietal,
TDLP; right parietal, TDRP; visceral, TDV.

Figure 4 (continued). Tentacle Nerve Fill Maps,
b) TN-filled from the right side, dorsal view.
Locator letters: left buccal, TDLB; right buccal,
TDRB; left cerebral, TDLC; right cerebral, TDRC;
left pedal, TVLPe; right pedal, TVRPe; left pleural,
TDLPl; right pleural, TDRPl; left parietal, TDLP;
right parietal, TDRP; visceral, TDV.

Figure 4 (continued). Tentacle Nerve Fill Maps,
c) TN-filled from the left side, ventral view.
Locator letters: left buccal, TVLB; right buccal,
TVRB; left cerebral, TVLC; right cerebral, TVRC;
left pedal, TDLPe; right pedal, TDRPe; left pleural,
TVLPl; right pleural, TVRPl; left parietal, TVLP;
right parietal, TVRP; visceral, TW.

Figure 4 (continued). Tentacle Nerve Fill Maps,
d) TN-filled from the right side, ventral view.
Locator letters: left buccal, TVLB; right buccal,
TVRB; left cerebral, TVLC; right cerebral, TVRC;
left pedal, TDLPe; right pedal, TDRPe; left pleural,
TVLPl; right pleural, TVRPl; left parietal, TVLP;
right parietal, TVRP; visceral, TW.

Figure 5. Superior Lip Nerve Fill Maps,
a) SLN-filled from the left side, dorsal view.
Locator letters: left buccal, SDLB; right buccal,
SDRB; left cerebral, SDLC; right cerebral, SDRC;
left pedal, SVLPe; right pedal, SVRPe; left pleural,
SDLPl; right pleural, SDRPl; left parietal, SDLP;
right parietal, SDRP; visceral, SDV.

Figure 5 (continued). Superior Lip Nerve Fill Maps,
b) SLN-filled from the right side, dorsal view.
Locator letters: left buccal, SDLB; right buccal,
SDRB; left cerebral, SDLC; right cerebral, SDRC;
left pedal, SVLPe; right pedal, SVRPe; left pleural,
SDLPl; right pleural, SDRPl; left parietal, SDLP; -
right parietal, SDRP; visceral, SDV.

Figure 5 (continued). Superior Lip Nerve Fill Maps,
b) SLN-filled from the right side, dorsal view.
Locator letters: left buccal, SDLB; right buccal,
SDRB; left cerebral, SDLC; right cerebral, SDRC;
left pedal, SVLPe; right pedal, SVRPe; left pleural,
SDLPl; right pleural, SDRPl; left parietal, SDLP; .
right parietal, SDRP; visceral, SDV.

Figure 5 (continued). Superior Lip Nerve Fill Maps,
c) SLN-filled from the left side, ventral view.
Locator letters: left buccal, SVLB; right buccal,
SVRB; left cerebral, SVLC; right cerebral, SVRC;
left pedal, SDLPe; right pedal, SDRPe; left pleural,
SVLPl; right pleural, SVRPl; left parietal, SVLP;
right parietal, SVRP; visceral, SW.

Figure 5 (continued). Superior Lip Nerve Fill Maps,
d) SLN-filled from the right side, ventral view.
Locator letters: left buccal, SVLB; right buccal,
SVRB; left cerebral, SVLC; right cerebral, SVRC;
left pedal, SDLPe; right pedal, SDRPe; left pleural,
SVLPl; right pleural, SVRPl; left parietal, SVLP;
right parietal, SVRP; visceral, SW. x: also
known as CV6 and CV7. y: also known as PRV3.

Figure 6. Median Lip Nerve Fill Maps,
a) MLN-filled from the left side, dorsal view.
Locator letters: left buccal, MDLB; right buccal,
MDRB; left cerebral, MDLC; right cerebral, MDRC;
left pedal, MVLPe; right pedal, MVRPe; left pleural,
MDLPl; right pleural, MDRPl; left parietal, MDLP;
right parietal, MDRP; visceral, MDV. x: also
known as VDl.

Figure 6 (continued). Median Lip Nerve Fill Maps,
b) MLN-filled from the right side, dorsal view.
Locator letters: left buccal, MDLB; right buccal,
MDRB; left cerebral, MDLC; right cerebral, MDRC;
left pedal, MVLPe; right pedal, MVRPe; left pleural,
MDLPl; right pleural, MDRPl; left parietal, MDLP;
right parietal, MDRP; visceral, MDV.

Figure 6 (continued). Median Lip Nerve Fill Maps,
c) MLN-filled from the left side, ventral view.
Locator letters: left buccal, MVLB; right buccal,
MVRB; left cerebral, MVLC; right cerebral, MVRC;
left pedal, MDLPe; right pedal, MDRPe; left pleural,
MVLPl; right pleural, MVRPl; left parietal, MVLP;
right parietal, MVRP; visceral, MW.

Figure 6 (continued). Median Lip Nerve Fill Maps,
d) MLN-filled from the right side, ventral view.
Locator letters: left buccal, MVLB; right buccal,
MVRB; left cerebral, MVLC; right cerebral, MVRC;
left pedal, MDLPe; right pedal, MDRPe; left pleural,
MVLPl; right pleural, MVRPl; left parietal, MVLP;
right parietal, MVRP; visceral, MW. x: also
known as PRV2. y: also known as W4.

(see Figures 6a and 6b). Right and left ganglia are
somewhat asymmetrical due to an extra lobe on the
right side from which the penis nerve originates.
This causes some shifting of cells, as well as addi-
tional cells in the right ganglion.
General Features
Table 1 describes pairs of individual cells
or clusters which were identified in fills of different
nerves but may be identical cells. This is suggested
by their location, size, shape, position relative to
other stained cells, and intraganglionic axons that
may stain consistently. These factors only suggest
similarity. Physiological investigation and intra-
cellular staining will be necessary to demonstrate
whether or not these cells are the same.
A few generalizations can be made about the
maps when they are observed as a whole. The one most
important characteristic to note is that most of the
cells that fill are not in the ganglion from which the
nerve originates. Only a small percentage of the
filled cells are in the cerebral ganglia. There are
also very definite trends in the sizes of cells that
fill. All of the individual cells that fill, with
only two exceptions, are large or medium in size.
Typically small cells only occur in groups. There are

Table 1
Cells From Different Maps That May Be the Same But
Could Not Be Positively Identified
Cell names
MDLPr5-M;sub and
Supporting remarks in favor of or
against cells being the same______
May be either the same cell, be-
cause of location, or two cells
that reside next to each other.
MDRPeQ-S and
TDV5-L;srf, and
These cells occur in the same place,
May be either the same cells or
cells in the same group.
These three cells occur in the same
place within the variation known
for individual cells, and are
nearly the same size. Only CDV5-L
is a sub cell, the others are srf
These cells may be part of MDRPd-K
because of their size and location.
a few cases where small cells are labelled as individ-
uals. This exception is made when the small cell in
question is consistently present in the fills and is
not part of a cluster, and particularly when a compar-
able larger cell is present on the opposite side. As
far as groups are concerned, all groups are either
small or medium cells, with one exception, a group of
large cells. More than half of the groups are small
cells. The patterns of groups, in terms of number and
positioning, are very consistent. The spatial rela-

tionship of cells in a group is also consistent from
one fill to the next. There is usually little varia-
tion in the number of cells in a group, within a
particular range. For example, MVLCA-M varied from
four to seven cells, never any more or less. Also,
many individuals have groups associated with them.
These groups are consistent in their number as well as
their spatial relationships in terms of the individual
cell. Often a group could be recognized by its
"shape" in the ganglion. For example, the cells of
CDRPec-S, in the right pedal ganglion on Figure 3d,
are in pairs, in a line. They occurred on every back-
fill in this configuration.
Soecial Cases
- -
Group A is an interesting group of cells in
the buccal ganglia. These cells are most prominent in
the ventral TN maps. In nearly all cases, the groups
in both the right and left ganglia fill regardless of
which TN was filled. The groups are easily identified
by their axons which connect the two groups through
the buccal connective. The cells occasionally reside
on the dorsal side (see TN Dorsal R and L maps, Figures
4a and 4b). Dorsal maps of the SLN show cells in a
similar position but without the complete connecting
axons. In this case these may be similar sized cells

in a similar place on the reverse side. Similar cells
were not seen in MLN backfills.
Two groups of cells show considerable varia-
bility in number. In group TDLCF-S two cells usually
fill, but in one case there were four cells and in one
case there were eight cells in this group. In group
TDLCa-S, six cells usually fill except in one case
where 18 cells filled. On the maps, groups like this
are represented by the number of cells that fill most
often. The group TDLPrE-S has two cells that appear
consistently. They seem to remain adjacent to each
other and move around within the group together. The
two cells are always the same size and shape and they
always stain to the same intensity. In group CVRPde-S
one cell is always located below and to the right of
the cells in group CVRPdb-L. It was grouped with
CVRPde because of its size, but may actually be asso-
ciated with the other cells. Another group of cells
to note is CDLPeZ-S and CVLPeZ-S. These two groups
may actually be one group which lie on the edge of the
ganglion and overlap onto the dorsal and/or ventral
side, thus creating the image of two groups as they
vary from one preparation to the next. The same
applies to the groups MVLPG-s and MDLPG-S.
In the pleural ganglion one group has an
unusual structure. MVLPlB-S (left and right fill),

MDLPlB-S (left and right fill) and MVRPlB-S (left and
right fill) are all split into two groups at the base
of the Pe-Pl connective. Each group is not totally
split, but usually two very distinct lobes of a large
group can be seen. In one case one group creeps over
the edge to the dorsal side as described above. Each
side of this group was therefore counted separately.
All of the cells identified are listed in
Table 2a, the individual cells, and 2b, the groups of
cells. (This table is located in Appendix A.) Each
table is organized with the cells in numerical or
alphabetical order, going through the SLN, MLN, TN,
and CBC maps. Each cell or group is given one line.
If the same cell fills from both the right and the
left it is still listed on the same line in the same
column (this applies only to individual cells, as the
same group filled from the right or left may have a
different number of cells). If there is a right and
a left cell or group on the same face (dorsal or
ventral) they share a line but are in different
columns. On Table 2b, medium groups immediately
preceed similar small groups. The cells which filled
from different nerves and are thought to be the same
(have the same number or letter), but were not con-
firmed, are grouped together for comparison.

There is some variability in the efficiency of
cobalt-backfilling. The metal ions do not diffuse
down all axons to the same extent every time, thus
not every cell fills every time. Usually, though,
cells fill consistently, and as a result appear to be
the same color from fill to fill. The cells that fill
in this manner were always included on the maps.
Occasionally some cells appeared light in color. The
cobalt ions may not travel down the process as well
for some reason, being either too small or not having
enough time, thus these cells do not fill as well.
Cells which were filled lightly, especially when they
filled inconsistently, were not included. To compen-
sate for these variations many preparations of each
type were done. Of course, some cells filled with
varying degrees between these two extremes. Each case
was considered individually as to whether or not it
belonged on the maps. Characteristics which prompted
the consideration of a given cell or group included

consistent filling, regular location, and similar
companion cells. Characteristics that prompted exclu-
sion were irregular filling frequency and cells that
usually filled lightly.
This method of identifying cells gives no
indication as to the physiological roles of the cells
being filled. There are four types of cells that
could be filled (refer to darkened cells in Figure 1
to see what the possible structural organizations of
these cells could be). First, there could be secondary
receptors which synapse with primary receptors located
in the epithelium. The only ones in this category that
would stain would be cells which had synapses outside
the CNS, probably just below the sensory epithelium.
Filled cells could also be primary receptors with
endings in the periphery. Although theoretically such
cells could be either chemoreceptors or mechanorecep-
tors, to date no primary chemoreceptors have been
identified in gastropod CNS, while several investi-
gators have reported central primary mechanoreceptors
(Audesirk, 1980) There could also be complex-type
neurons or intermediate neurons which function as both
primary and secondary receptors depending on whether
they are receiving or integrating information
(Audesirk, 1980). Finally, cells could be motor

neurons responsible for movement of appropriate
An interesting observation concerns the
identification of the tentacle nerve. Since it is
usually attached to the optic nerve it is often
difficult to distinguish the two. It was discovered
that if the eye of the snail is squeezed prior to
removing the brain, dark fluid from the eye fills the
optic nerve and the tentacle nerve remains unstained
and positively identified.
Other Studies
Several papers have been published which iden-
tify particular cells on the basis of location and
electrophysiological properties (see Table 3 for
references). Some of these cells can be matched
exactly with cells represented here on the maps.
These are indicated on Table 3. Other cells may be
represented here but positive identification awaits
electrophysiological studies. These probable matches,
indicated in Table 3, are based on location, size,
and shape. One example of a probable match is from
cells examined in the study recently done by McCrohan
(1984). Her CV6 and CV7 cells correspond to the cells
grouped together as SVLCC-M in this paper. The com-
parison was easily made because of anatomical descrip-

Table 3
Comparison of Cells in This and Other Studies
Cells Identified Here Cells In Other Studies Reference
SVLCC-M CV6 and CV7 McCrohan 1984
CDLB7-M;srf 1 Goldschmeding
CVRBK-M & S CVRBL-S g-region cells m-region cells 1977
CVLCd-M & CVRCd-M Identified but no name
CVLC16-M & CVRC16-M Category 1, not named
SDVC-M MDV5-L;srf F group VDl Benjamin 1981
LVRP3-M;srf (SLN map) PRV3 Benjamin 1972
LVRP3-M;srf (MLN map) PRV2
MWl-M; srf W4
MVLPlB-M D group Haydon 1982
CDV5-L;sub CDRPD-M A10 PI Kiss 1979
Cells That Are Probable Matches But -Lack
Conclusive Evidence
CDLB7-M;srf BL4C1 Rose 1981b
Type of cells VanSwigchem
classified 1981
as LYC
MVRP3-L;srf &

tions in McCrohan's paper. Another study of Lymnaea
was also compared to the maps here. Lever (1977)
studied neurons involved in tentacle contraction. Due
to differences in orientation and the lack of identi-
fication of discrete clusters by Lever, certain
identification was not possible.
Overall, this study encompasses most of the
cells that have processes innervating the major sensory
organs, the lips and the tentacles, and thus have a
role in sensory reception and output. These maps are
the starting point for verifying the cell's identity
and possibly their roles, via electrophysiological
investigations. As a continuation of this, the second
part of the study identifies peripheral elements, the
receptors which perceive external stimuli.

The purpose of this part of the study is to
identify the elements involved in stimulus reception
and to suggest how they may be connected to each
other and to the CNS. More specifically, neurons
contacting the environment (primary receptors) will be
identified, along with their organization in the
epithelium and their distribution on the face of the
snail. These neurons are the peripheral element of
the chemosensory circuitry. They are the neurons
which receive information and relay it to the cells
in the CNS which were identified in Part I. The end
of this study is a cobalt backfilling study. Filling
the MLN toward the periphery suggests some patterns
of organization of the circuitry between the peripheral
elements and the CNS.
Vertebrate vs. Invertebrate Sensory Epithelium
Although there are many similarities between
some of the individual components of vertebrate olfac-
tory epithelium and the epithelium on the lip of
Lymnaea, as a whole the two tissues are quite different.

Vertebrate olfactory epithelium has one primary
function, to detect odorant molecules. This function
is reflected in a relatively noncomplex structure. One
type of neuron, in one or more stages of development,
is interspersed throughout epithelial cells. The
neurons are localized to a specialized region of the
nasal cavity epithelium (Frisch, 1967). On the other
hand, Lymnaea sensory epithelium is quite complex.
Along with functioning as a body covering, the skin of
the entire snail is sensitive, to various extents, to
a wide variety of stimuli, including chemical, mechan-
ical, and light (Bovjerg, 1968; Stoll, 1976; Lever,
1977). All of the receptors for these stimuli are
located in the skin, and usually are localized to
specialized areas. With so many types of stimuli being
received by so many different receptors, it is not
surprising to see such a complex structure, packed with
cell types that are difficult to distinguish from one
another. This complex structure is common in many
invertebrates such as Helix pomatia (Benedeczky, 1977;
Wondrak, 1981), Aplysia (Emery and Audesirk, 1978),
Arion ater (Wright, 1974), Nassarius reticulatus
(Crisp, 1971), Pleurobranchaea californica (Davis and
Matera, 1982), and also crustaca (Ache, 1982).

Adult snails of the species Lymnaea stagnalis,
maintained as described in Part I, were used for this
study. Specimens averaging 1.9g were selected. The
lip (see Figure 7) was dissected in Ringers and proc-
essed according to one of the following methods.
Light and electron microscope preparations. Tis-
sues were fixed in Karnovsky's (Karnovsky, 1968) over-
night, rinsed in Ringers, postfixed in osmium (1%) for 2-
3 hours, dehydrated in an acetone series, infiltrated in
acetone/Epon-Adralite resin solutions (Mokotoff, 1978)
for 2-3 days in a dessicator, and embedded in Epon-Adra-
lite. Section's for the light microscope were cut at
0.20-0.25pm and stained with Richardson's or Toludine
Blue (Humanson, 1979). Sections for the transmission~EM
were cut at 0.050-0.075pm and stained with uranyl acetate
and lead citrate (Mokotoff, 1978). High voltage EM sec-
tions were cut at 0.20-0.25pm and stained with uranyl
acetate and triple lead stain (Sato, 1968).
Reconstruction. The whole head was fixed in
calcium formalin (Humanson, 1979), dehydrated in an

alcohol series, embedded, and sectioned at 5pm.
Tissues were stained with Mallory's triple stain
(Humason, 1979) and mounted in permount on glass
slides. Each sample, cut at a cross section, and
sampled at regular intervals (one section every 40pm),
was drawn, marking the location of epithelial regions.
These were then drawn as a ventral face drawing. The
area was estimated by a point counting method (Under-
wood, 1970). The point counting method utilizes a
grid on a transparent sheet which is placed on top of
the photograph. The intersections of the horizontal
and vertical lines (the points) are counted. The
ratio of the number of points in the area to be esti-
mated over the total number of points is the estimation
of the area. Error is reduced by multiple calculations.
Cobalt backfilling. The cobalt technique
described in Part I was used with the following
changes. The lip area was dissected from the rest of
the snail, leaving the MLN intact and cut to a long
length. Only the MLN was filled, in the direction of
the lip (as opposed to toward the CNS, as was done in
Part I). Preparations were left to diffuse for only
17-19 hours, to reduce tissue deterioration, and the
precipitate was developed with 14% (v/v) ammonium
sulfide. During the diffusion tissues were bathed in

Ringers with 0.15% (w/v) penicillin in solution.
Tissues were then prepared as described for LM and EM

Overall Morphology of Head Skin
This study revealed three regions of epithelium
on the lip of Lymnaea with distinct characteristics.
The reconstruction from serial sections gives an idea
of the distribution of these three regions on the
ventral side of the snail's face (Figures 7 and 8).
The first and most important is the ganglion rich
region (GRR), (Figures 8 and 9-13). The GRR surrounds
the mouth and extends down the center of the oral
lappet. It is defined by the presence of numerous
ganglia in the subepithelium, adjacent to the epithe-
lial cells. These ganglia are generally roundish, but
can vary in shape, and can be as large as 30jom x 40;um.
The epithelial cells are tall, columnar, though not
regular in shape, and are the tallest of any of the
regions, 28-36jum, and are densely packed together, i.e.,
there are more cells in a given unit of area in this
region than there are in any of the other regions.
This region is the area that has the most sensory
receptors, and therefore is the region that is most

Figure 7. Dorsal and Ventral Views of Lymnaea Head,
a) Dorsal view of head of Lymnaea. Tent., tentacle;
O.L., oral lappet.
b) Ventral view of head of Lymnaea showing arrangement
of the 3 regions of epithelium. See text for detailed
description of these areas. The sensory groove appar-
ently follows the ganglion-rich region from the edge of
the mouth out to the edge of the oral lappet (see
arrows along GRR on right side). PR, plain region;
MR, mucous region.

Figure 8. Epithelial Regions.
A low magnification picture of a cross section
through the lip. Illustrates the relative arrange-
ment of the three epithelial regions and the
sensory groove. Mag. 150x.
The following abbreviations are used in the photo-
graphs and illustrations: BB, basal body; BC,
basal cell; C, cilia; CMV, curly microvilli; DV,
disjoined neuron; Epith., epithelium; G, ganglion;
GC/ ganglion cell; GRR, ganglion-rich region; M,
mitochondria; MC, motile cilia; MR, mucous region;
Mu, mucous c-ell; Mv, microvilli; N, neuron cell
nucleus; Ne, nerve; n, epithelial cell nucleus;
PR, plain region; R, roots of cilia; SG, sensory
groove; SMV, straight microvilli.

Figure 9. Ganglion-Rich Region (GRR) of Lip
Density and packing of cells is easily seen. Arrows
indicate E4 process which extends through the entire
of epithelium. +, bundle of processes leaving
the epithelium; *, pieces of processes which wind
through plane of section. Mag. 2500x.

Figure 10. Ganglion-Rich Region of Lip Epithelium.
Adjacent plane of the same section as in Figure 9.
Arrows show portions of receptor endings passing in
and out of the section on the way down to a ganglion.
Mag. 250Ox.

Figure 11. Ganglion-Rich Region Near Groove.
Arrows indicate portions of processes in the sub-
epithelium and in the ganglion. Mag. 2500x.

Figure 12. Ganglion Processes in Lip Epithelium.
The area where processes enter the epithelium.
Arrows indicate process(es) traveling from the
ganglion to the epithelium. Mag. 2500x.

Figure 13. Receptor Ending E6.
In the ganglion-rich region. Mag. 8700x.

important in terms of chemoreception. Other inter-
esting characteristics of the GRR are as follows. The
basal ends of the epithelial cells are very irregular,
making it extremely difficult to distinguish one end
from the next. The ganglia in the subepithelium are
surrounded by a large number of connective tissue
fibers and a few pigment cells (Figure 11). Muscle
cells run along parallel with the base of the epithe-
lium. Occasionally processes from the ganglion cells
can be seen running along the base of the epithelium.
The subepithelial neurons have various shapes
in irregularly shaped, roundish ganglia. The nuclei,
the most prominent feature, are round to oblong with
well defined nuclear envelopes and obvious chromatin
alqng the edges and throughout the nuclei. The
nucleoplasm appears granular. Occasionally a part of
a process can be seen among the ganglion cells. In
some cases the ganglia lie in direct contact with the
epithelial cells, with nothing between them.
Another characteristic of the GRR is the
presence of curly microvilli (CMV) as well as straight
microvilli (SMV) (Figures 10 and 12). SMV are micro-
villi that appear similar to typical microvilli found
on any absorptive surface. They are short and stand
upright. CMV, on the other hand, are about only half
the width of SMV and are so densely interwound between

the SMV that they form a "spongy" layer. The SMV rise
to a height of 2.7-4.7jum, whereas the CMV are only
1.5-2.2/jm. CMV are found on all nonsensory cells and
on one sensory cell type and are found only in this
The second region of epithelium is the plain
region (PR) (Figures 14 and 15). It is defined as
Figure 14. Plain Region of Lip Epithelium.
Notice that there are no CMV, and that the cells
are more regular in appearance. Note similarities
in the nuclei of the E2 process at the and the
disjoined neuron. Mag. 2400x.
epithelium on a sensory structure which lacks ganglia
and lacks an appreciable number of mucous cells. The

Figure 15. Plain Region of Lip Epithelium.
Compare the nuclei of the epithelial cells to the
nuclei of the DN. Mag. 2800x.
cells in general are more regular in profile and are
not as dense as those in the GRR. There are a small
number of sensory endings which are attached to indi-
vidual cell bodies in the subepithelium. Overall the
cells are columnar, 26-29jum tall. All cells have only
SMV which are straight and 1.9-2.7/am tall. The cells
are more uniform in size and shape than those in the
GRR, and are not as dense. The subepithelium in this
area has lots of fibers, muscle, pigment cells, and

some disjoined neuron cell bodies. (Disjoined cell
bodies are neuron cells like those in the ganglia,
which are not associated with a ganglion.)
The third region is the mucous region (MR)
(Figure 16). It is defined by abundant mucous cell
Figure 16. Mucous Region of Lip Epithelium.
Note the large mucous producing glands and cells.
Also the regularity and consistent shape of the
cells. E3 is shown in association with a DN.
Notice the lack of ganglia and appreciable number
of DN. Mag. 2700x.
and the lack of ganglia. The types of mucous cells
are described by Zylstra (1972b) and will not be
discussed here. The epithelial cells in general
resemble those of the plain region and are the same

size. The subepithelium also contains abundant fibers
and muscle tissue.
A fourth region, not found on the lip, is un-
specialized epithelium (UE). It is located on the
nonspecialized skin of the rest of the body and is
used as a control for this study (Figure 17). It was
Figure 17. Unspecialized Epithelium, From Head Skin.
Cells are very regular in size and stature. Basal
lamina and basal edge of epithelial layer are well
defined. +, another receptor process. Mag. 3400x.
sampled from an area on the head, away from the eyes,
tentacles, and lip for this study. It is characterized
by its uniformity and medium height columnar cells
(23-25jum) which have microvilli, though not as regularly

structured as those in the lip regions. Only a very
small number of mucous cells and sensory endings are
found in this region.
Although these regions are well defined, the
transition between them is not abrupt. Individual
neuron cell bodies could extend into the other regions,
or a mucous gland could occur in an adjacent region
near the boundary. These overlaps are small and do not
interfere with region distinction.
Distribution of Epithelial Regions of the Lip
A reconstruction of a typical snail gave a
general idea of the arrangement of the three regions.
This reconstruction suggests the arrangement in
Figure 7b. The GRR surrounds the mouth and extends
down the oral lappet, apparently following a sensory
groove similar to the one described by Bicker (1982)
in Pleurobranchea californicia. This groove is filled
with very long cilia attached to the epithelial cells
in the groove, (Figure 10), and in the electron micro-
graphs is seen associated with the ganglia. These
cilia, which are extremely long, are hypothesized to
be motile cilia, cilia which are used to create water
currents over the surface, and are thought to have no
sensory function (Emery and Audesirk, 1978). The
reconstruction suggests that approximately 15% of the

surface areas is GRR, 13% is PR, and the remaining
72% is MR. Observations of less elaborate samples
follow a similar pattern.
Criteria for Identification of Receptors
There are five criteria that a cell may possess
in order to be considered a receptor, whether the
nucleus is in the epithelium or in the subepithelium.
These characteristics, when considered all together,
apply only to the neurons which form sensory receptors,
but a nonsensory cell may possess one or another
characteristic occasionally. The more characteristics
a cell has the stronger the evidence for it being a
receptor. Most of the receptors described have at
least four of these characteristics. Reflecting back
on Figure 1, the cells that are being considered are
neurons which either have the cell body and nucleus in
the subepithelium and a process (referred to as a
receptor process) in the epithelial layer, or have the
entire cell, nucleus and specialized ending, in the
epithelial layer. The latter case is then connected
via some process to the neurons in the subepithelium.
a) The first factor is based on structure.
A neuron with its cell body in a ganglion in the sub-
epithelium which sends a specialized process through
the epithelium, having its apical end exposed to the

environment, is a typical structure for receptors.
(See arrows on Figures 9 and 11.) It is the structure
seen most frequently in gastropods. A cell with this
structure is indeed a neuron, and most likely a sensory
receptor. An alternate structure is an intraepithelial
cell with a specialized apical region and some connec-
tion to the ganglia or disjoined neurons in the sub-
epithelium. Although this structure is uncommon in
invertebrates, it has been identified in Pleuro-
branchaea californica (Matera and Davis, 1982) and
resembles the structure seen in vertebrate olfactory
epithelium (Frisch, 1967; Graziadei and Monti Graziadei,
1978). Cells that do not fit these criteria are non-
sensory epithelial cells, which do not have any
connection to any neuron-type structures (i.e., ganglia
and disjoined neurons).
b) The second factor is the morphological
appearance of the nucleus. The nuclei of the neurons
in the ganglia have distinct heterochromatin around the
edges and granular nucleoplasm (Figure 11). These can
be contrasted to the nuclei of unspecialized epithelial
cells (Figure 17) and to nonneuron epithelial cells of
the lip (Figure 15). The latter, in particular, have
indistinct chromatin and are less granular. This
criterion is used when the cell body in question is not

part of the ganglia and therefore its identity is
c) The third factor is location. The
receptors considered were found primarily on the lips,
and infrequently on unspecialized epithelium. On the
lip, receptors are further localized primarily, though
not exclusively, to the GRR. This region was previously
hypothesized, based on other morphological evidence, to
be the most important sensory region. Thus, cells which
occur in this area have a large probability of being
sensory receptors. Since there has to be a physical
connection between a receptor and the next neuron in
the circuitry, it follows that the frequency of the two
elements should parallel each other and their physical
proximity should be the same. This comparison is made
between elements in the epithelium, sensory processes,
and intraepithelial sensory cells, and elements in the
subepithelium, ganglia cells and disjoined neurons. In
the GRR where there are lots of neuron cell bodies in
the subepithelium, there are also lots of processes of
cells in the epithelium. This may seem like a moot
point but when considering other areas it becomes more
clear. In the PR there are fewer disjoined neuron cell
bodies in the subepithelium (remember there are no
ganglia here or in the rest of the regions) and an
equal reduction in the number of processes and intra-

epithelial cells in the epithelium. More importantly
though, the two elements are more distinctly seen in
the same physical proximity. Where a neuron cell body
exists in the subepithelium, there are processes or
intraepithelial cells in the epithelium. This is even
more apparent in the MR where the numbers are even
smaller and quite obvious in the UE where the elements
(receptor cells, basal cells, and disjoined neurons)
are rare. The parallel occurrence of a neuron in the
subepithelium with a process or intraepithelial
receptor is strong evidence that the element in the
epithelium is indeed a sensory receptor.
d) The fourth factor is the presence of cilia.
While not found on all receptor types, cilia are a
characteristic of many molluscan receptors and of
vertebrate olfactory receptors (Frisch, 1967). Cilia
are believed to function as a mechanism of increasing
surface area for transduction. Based on this, they
should be, and are, most readily found where the cilia
are most easily contacted by the stimuli. An exception
to this is the motile cilia which are used to create
water currents to move water over the surface of the
skin. A variation of this theme is microvilli which
may have been derived for special use by sensory
receptors. They appear morphologically different from
the microvilli which are classified here as SMV or CMV,

often being longer and thinner. (Refer back to the
descriptions of SMV and CMV earlier in the results.)
e) The fifth factor is the receptor's similar-
ity to receptors identified in other studies. These
can be from both gastropods as well as vertebrates. In
this section each receptor described will be compared
to similar receptors for the purpose of identification,
and they will then be compared in detail in the discus-
These features were assessed using a number of
techniques. Thick sections provide a greater possibil-
ity of viewing an entire process in one section.
Serial sections were used to establish complete
connections between cells or parts of a single cell, as
well as the number of cilia or microvilli, and presence
or lack of nuclei.
Previously Unidentified Receptors
Characteristic drawings of each cell type are
presented in Figure 18. Table 4 summarizes descriptions
and locations.
Ending type 1 (El) is a bulb cell with CMV, and
in an intraepithelial receptor (Figure 9). It has a
slightly enlarged ending covered with both SMV and CMV.
It is 1.2-2.0jum at the widest point. It often has
large vacuoles near the apical end. The cell tapers

Figure 18. Characteristic Drawings of Receptors.
Drawings of the apical ends of the receptors found
in the epithelium. Also illustrates a basal cell

Table 4
Cells in Lymnaea Lip Epithelium
Cells in Lymnaea Lip Epithelium Apical
Notable Characteristics Structures Location
Cell Label Previously Unidentified Receptors SMV CMV Cilia GRR PR MR UE
Type 1 El Bulb-shaped ending yes yes no yes yes yes no
Type 2 E2 Bulb-shaped ending yes no no no* yes yes no
Type 3 E3 Thin subepithelial process, wide ending yes no no ? yes yes no
Cells in Lymnaea Lip Epithelium Notable Characteristics Apical Structures Locations
Cell Label Previously Identified Receptors SMV CMV Cilia GRR PR MR UE
Type 4 E4 Cup-like pit holding cilia no no 1-3 yes yes yes yes
Type 5 E5 Cup-like pit holding cilia yes no 5+ yes yes no no
Type 6 E6 Microvilli in irregular array yes no no yes no no yes
Basal BC At basal edge of epithelial layer - - - yes yes yes yes
* If El is a subtype of E2 then this cell type is in the GRR,
See Discussion for a more detailed explanation.
but not in this form.

to a very thin basal end and has an elongated nucleus
in the middle of the cell. Receptor criteria: a) El
is an intraepithelial cell connected via a thin process
to either ganglion cells or disjoined neurons in the
subepithelium, b) The nucleus shows distinct hetero-
chromatin and granular nucleoplasm, which is very
similar to those of other receptors, ganglion cells,
and vertebrate olfactory epithelial receptors (Monti
Graziadei, 1979). c) El is found only in the GRR
where its frequency is one out of every two or three
cells. It is the most abundant cell type in the GRR.
d) This receptor does not have cilia, but it has SMV
and CMV. (See the section in the Discussion on CMV.)
e) Intraepithelial receptors have usually been
thought to be rare in gastropods, but have been
described in a few cases. Davis and Matera (1982)
describe a similar intraepithelial receptor as being
very abundant in Pleurobranchaea California. El is
also similar to the type 4 receptor described by
Wright (1974) for Arion ater, also an intraepithelial
Ending type 2 (E2) is a bulb cell without CMV,
and otherwise is identical to El (Figure 19). It has
all of the same characteristics of El. Receptor
criteria: a) E2, like El, is connected to ganglia
cells or disjoined neurons via a thin process of the

Figure 19. Receptor Ending E2.
E2 is shown in the plain region, connected to a
disjoined neuron, via a small process, see arrow.
This connection is the same as that seen for El in
the GRR, a much more difficult connection to find
in micrographs. Mag. 40Ox.
cell. (See comparison of E2 and El in Figure 18.)
b) The nucleus of E2 is identical to those of El,
neurons in subepithelial ganglia, and vertebrate
olfactory receptors. c) E2 is found only in the MR
and the PR, always in association with disjoined
neurons where there are no ganglia. d) E2 does not
have cilia, only SMV. e) The similarity of E2 to
other cells is the same as El, again with the presence
of only SMV.
Ending type 3 (E3) is also an intraepithelial
receptor. It is a medium width cell, 2.1-2.6jum wide,

(Figure 16). Receptor criteria: a) E3 is connected
via a thin process which extends from the basal end of
the cell, through the subepithelium, to a disjoined
neuron cell body, b) The nucleus of E3 is identical
to El, E2, subepithelial neurons in ganglia, disjoined
neurons, and vertebrate olfactory receptors. c) E3
is easily identified in the PR and the MR, always
connected to or near a disjoined neuron cell body.
Because of the density of the cells in the GRR, and
their elongated shapes, it is not possible to tell if
E3 is in this region. d) E3 has only SMV at its
apical surface. There are no cilia. e) No other
receptor previously identified appears to be like E3.
Receptors Previously Identified
Ending type 4 (E4) is a process with an apical
cup-like pit containing one to four cilia (Figure 9).
It is a narrow process, 0.1-0.2jum in diameter. The
cytoplasm is generally electron lucent, is filled with
filaments, and contains several long mitochondria near
the apical end. Receptor criteria: a) The process of
E4 extends from the cell body in the ganglion, through
the epithelium, to the surface. b) The comparison of
nuclei does not apply here because the basis of com-
parison is based on the ganglion cells, which these
processes come from, c) E4 is found in all regions

of the epithelium, GRR, PR, MR, and UE. The frequency
of E4 in each region declines in proportion to the
decline of neuron cell bodies in the subepithelium of
each region. In the GRR the frequency is one out of
every four or five cells. In the PR, MR, and the UE,
E4 is always found near a disjoined neuron when no
direct connection can be seen. d) E4 has one to
four cilia in the cup-like pit. The cilia have basal
bodies from which roots extend, which may be as long
as the epithelial layer itself. There are no micro-
villi of any kind. e) The seemingly ubiquitous
occurrence of El in every gastropod that was considered
for comparison is a strong indication that this cell
is indeed a receptor. In many of these studies, E4
is the most common receptor for that species. E4 is
most likely the same receptor as Zylstra's (1972a)
type 3 and Zaitseva's (1981) type 1, both for Lymnaea
stagnalis. Emery and Audesirk's (1978) type 1 of
Aplysia californicia and Aplysia brasiliana and
Wright's (1974) type 1 of Arion ater are also nearly
identical to E4.
Ending type 5 (E5) is very similar to E4, but
not exactly the same (Figure 14). E5 has more cilia,
(five or more), and also has SMV around the periphery
of the cup-like pit. In other aspects E5 is just like
E4. Receptor criteria: a) E5 has a process which

also extends from the ganglion cells, through the
epithelium, to the surface, b) The same situation
applies as does for E4. c) E5 is found most frequently
in the GRR, and to a lesser extent in the PR. d) E5
has five or more cilia in the cup-like pit. These
cilia have basal bodies and roots that may be as long
as the epithelial layer. Along with the cilia there
are SMV on the apical surface surrounding the pit, but
not in it. e) E5 may be the same receptor as those
identified by two other investigators. Zylstra's
(1972a) types 4 and 5 of Lymnaea are very similar to
E5, although they do not seem to match exactly. Also,
Benedeczky's (1977) receptor type with cilia and micro-
villi of Helix pomatia is nearly identical to E5.
Ending type 6 (E6) has a slightly enlarged
ending, 1.0-1.5jum wide (Figure 13). The cytoplasm is
electron lucent and is packed with mitochondria.
Receptor criteria: a) The process of E6, like E4
and E5, can be seen extending from neurons in sub-
epithelial ganglia, through the epithelium, to the
surface, b) The same situation applies as does for
E4 and E5. c) E6 occurs infrequently in both the
GRR and the UE. d) E6 does not have cilia, but
rather microvilli which do not appear to be like SMV
or CMV. These microvilli, as they are referred to by
other investigators (Zylstra, 1972a), are very similar

in appearance to cilia on vertebrate olfactory receptors
(Frisch, 1967). Both are attached to the surface ran-
domly, are seemingly non-rigid, and lie in the mucous
layer. e) E6 is probably the same receptor as that
identified by Zylstra (1972a) in Lymnaea as type 1.
E6 also, except for the lack of basal feet and intra-
epithelial nuclei, looks almost identical to vertebrate
olfactory receptors (Frisch, 1967) .
Basal Cell
Basal cells are cells which are located in the
epithelium, at or near the basal region (Figures 9, 14,
16, and 17). They do not appear to have any processes
extending toward the surface or extending into the
subepithelium. They occur infrequently, but are found
in all four regions. The most distinguishing feature
is the large nucleus, which nearly fills the cytoplasm.
It has a very dark nuclear envelope and large chromatin
spots. The nucleoplasm is medium in electron density
and is very granular. They appear identical to ganglia
cells in all aspects including shape, size, and
staining characteristics. The basal cell is not
considered to be a receptor, but is considered to have
an important role in reception because of several
factors. They are found in all regions, but have a
declining frequency which parallels that of receptors

in general, i.e., numerous in the GRR, less in the PR,
even less in the MR, and only occasionally in the UE.
Their location in the epithelial layer is obvious,
particularly in the areas with few receptors. They are
always found near a receptor or receptor ending. This
is particularly noticeable in the UE where the recep-
tors are few and far between. The nucleus of these
cells, as described above, is identical to the nuclei
in the ganglion cells, intraepithelial receptors, and
vertebrate olfactory receptors (Monti Grazizdei, 1979).
No comparisons can be made to other cells described for
gastropods because there are no similar cells
described elsewhere. They do appear to be very similar
to basal cells described in vertebrate olfactory
epithelium, both in morphology and location.
Unspecialized Epithelium
Unspecialized epithelium comes from an area
which is not specialized (Figure 17). The sample for
this study was taken from the head, away from the lips
and tentacles. Regular features of this region include
the following. The sides of the individual cells
interdigitate with each other for the entire length of
the cells. The nuclei, which are situated in the
basal to mid region of the cells, are oblong, have
nucleoli, and have chromatin in the center of the

nuclei. The nuclear envelope appears as a thin line
and the nucleoplasm is granular. The cytoplasm is
predominantly filled with mitochondria and dark stain-
ing vesicles. The basal end of the epithelium is
uniform but interdigitates into the subepithelial
layer. The subepithelium is filled with muscle
bundles, fibers, and very few disjoined neuron cell
The receptor endings that are found in this
area are either above or near a disjoined neuron and
a basal cell. The receptors that are found are E4
and E6. Other apparent receptor endings were found
in this area but they were not encompassed by this
Cobalt Backfilling
Cobalt backfilling, unfortunately, causes
tremendous destruction of the tissues that in infil-
trates. This limits the amount of information that
can be gained from the procedure. Figure 20 shows a
section of epithelium from the control. It shows very
little difference from the previous pictures, except
for more numerous patches of degradation. Figures 21
and 22 show cobalt backfilled epithelium. The apical
surface can generally only be distinguished by the
presence of cilia. Only general characteristics of

Figure 20. Cobalt Backfill Control.
GRR tissue was backfilled with distilled water.
No extreme morphological destruction. Arrows
show process of ganglion cell entering the
epithelium- Mag. 2500x.

Figure 21. Cobalt Backfilled GRR Epithelium.
Cobalt can be seen as very dark cytoplasm (see
arrows). Identities of cells and process types
in the epithelium are not easily determined
because of the extreme destruction of morpho-
logical features. Notice that not all of the
ganglia cells have cobalt in their cytoplasm.
Mag. 2500x.

Figure 22. Light Microscope Photograph of Cobalt
Filled Epithelium.
Shows cobalt throughout the tissue. It can be seen
in nerve trunks, ganglia, and areas in between.
(Cobalt is difficult to see at this magnification.)
The amount of destruction to epithelial cells is
obvious. Mag. 40Ox.

the tissues can be distinguished. Individual epithe-
lial cells can usually be determined, but not
specifically identified. Ganglia, ganglion cells, and
other large subepithelial structures can be identified.
In some of these structures cobalt deposits can be
seen. The cobalt can occur as clusters of extremely
dense globules or it can be more evenly dispersed
through a cell. Figure 21 shows that most, but not
all, ganglia cells fills with cobalt. Some of the
dendritic processes (probably E4, E5, or E6) can be
seen filled with cobalt. A couple of cases were found
(Figure 21) where some epithelial cells with nuclei
appear to be filled with cobalt. From these results
a proposal for the circuitry can be made (see

Most of the other studies on epidermal receptors
define structures as being receptors based on the
observations that they extend from subepidermal neural
structures, particularly ganglia (Emery and Audesirk,
1978; Zylstra, 1972a; Roubos, 1982). This is the
primary and often the only criteria for defining
receptors. From this point, nearly every study differs
in the way they classify the receptors. These classi-
fications are usually based on the differences in
structure of the apical end of the dendrites which
extend into the epithelium. Some studies use other
criteria for defining receptors. Some of these
include being stained by a stain specific for nervous
tissue, location, or general appearance.
The degree with which receptors are depicted as
entire entities varies widely. In Emery and Audesirk
(1978) Aplysia receptors are often pictured beginning
in the ganglia and extending through the entire layer
of the epithelium. Other studies, such as Roubos
(1982) and Zylstra (1972a), do not illustrate complete
connections with entire epithelial processes, but

rather show bundles of processes extending into the
epithelium or explain that the structural arrangement
was discerned from serial sections. This may not
necessarily be viewed as a lack of concrete evidence,
but rather indicates the irregularity of the tissue,
and perhaps the success of the sectioner. Occasionally
only indirect evidence is used to suggest that a
structure may be some type of receptor. Matera and
Davis (1982) indicate that the paddle cilia of Pleuro-
branchaea californica may be receptors based only on
their distribution in areas that have been shown
behaviorally to be chemosensory.
Another feature almost always mentioned as an
observation, but not a criterion for identification,
is the electron lucent cytoplasm of the dendritic
process in the epithelium. This characteristic,
though, is not inclusive of all receptor ending types.
Overall, the biggest piece of morphological
evidence for a structure being a receptor is its
connection to nervous tissue in the subepithelium.
This same piece of evidence applies here, also being
the most important criterion. No other study, though,
provides as many further criteria for defining recep-
tors as this one does, particularly for the receptors
which were previously unidentified.

Previously Unidentified Receptors
Previously unidentified receptors are cells
which have not been singled out as receptors for
Lymnaea in the past. Two of these, El and E2, are
nearly identical to each other. Their only morpho-
logical difference is the presence of CMV on El.
Since, though, CMV are particular to the GRR, then
El may just be a subtype of E2. This is more than
likely the case, since it appears that CMV is a
characteristic of the GRR only, and is not isolated
to this receptor.
El and E2 do resemble receptor type 4 of
Arion ater (Wright, 1974) and type 1 of Helix Pomatia
(Benedeczky, 1977). These receptors, though, were not
described as intraepithelial receptors. On the other
hand there are several intraepithelial receptors
described for other gastropods. Davis and Matera
(1982) describe an intraepithelial receptor which is
very abundant in Pleurobranchaea californica. Emery
and Audesirk (1978) also describe an intraepithelial
receptor in Aplysia. Both of these receptors are
E3 is a very different type of receptor from
anything described previously. In general terms it is
similar to the receptors described with microvilli
only, but it really does not resemble any other

receptor in any other way. Part of the reason why it
has not been described before may be that it is usually
not found in the GRR. It either may not be there
because its function requires it to be elsewhere, or,
more likely, it is just difficult to see there.
Because of the extreme density of cells in the GRR,
the tight packing, the elongation of the cells, and
the way the receptor processes twist around the cells,
it may not be possible to recognize it (thus the
question mark in Table 4).
Comparison to Other Studies
The rest of the receptors discussed here were
previously identified in other studies by the structure
of their apical ends. In almost all of these, recep-
tors were classified in either one of two ways, or a
combination of both. The most popular method was to
categorize cells by the type and number of protrusions
(cilia and microvilli) on the apical end. The second
was categorizing endings by detailed structure of a
number of cellular components. While a combination of
these methods seems to produce the best results, this
study suggests that a large number of divisions may
not be necessary, particularly where function is
considered. This study attempted to include a more
complete characterization of the entire cell, as well

as other characteristics of the sensory epithelium.
Direct correlation to several other papers can be
Zylstra's (1972a) study of sensory regions of
Lymnaea was based on the number of cilia and the
structure of their roots. E4 and E6 are probably the
same as Zylstra's types 3 and 1 respectively. E5 also
resembles Zylstra's types 4 and 5, but they are not
completely identical. The differences lie in the
number of cilia, the cup-like pit, and the length of
the roots. There is insufficient evidence to discern
whether or not these are actually the same receptor.
Some of the distributions of cell types and frequencies
of cells within the lip region were different from
those noted by Zylstra. (No actual numbers were
available for comparison.)
Another study done on receptors of Lymnaea
was by Zaitseva (1981), who made no distinction as to
which areas of epithelium on the head were being
described. E4 is similar to Zaitseva's type 1. No
other types match but there are several similar
structural aspects. The winding and branching patterns
and ultrastructure descriptions are similar.
Benedeczky (1977) describes only two types of
receptors in the lip region of Helix pomatia. These
two cell types are categorized by the presence of

microvilli only, or microvilli and cilia, correlate
with E2 and E4 respectively. The apical and cellular
characteristics of the receptors from the two species
are very similar. No mention of structures similar to
CMV were made.
Emery and Audesirk (1978) also described two
types of sensory cells in Aplysia californicia and
Aplysia brasiliana. Here again, E4 of this study is
similar to their type 1, both having few cilia, as well
as other similar characteristics.
Studies of other gastropods have some aspects
of their sensory endings similar to those described
above. Direct correlations, though, are often not
possible. Studies done by Wondrak (1981) on the
ommatophore of Helix pomatia and Wright (1974) on
tentacles of Arion ater point out receptors that share
characteristics similar to these receptors. These are
the presence or lack of both cilia and microvilli, and
their relative arrangements. In particular, Wright's
type 1 is nearly identical to E4 and his type 4 is
similar to El.
One note, that was already mentioned, is the
seemingly ubiquitous occurrence of E4 in every gastro-
pod referred to. Because of their consistent occurrence
they probably have some uniform role or function in

the respective sensory systems. This will be discussed
in the section on function.
Basal Cells
Another feature of Lymnaea epithelium not
previously described is the basal cells. They are
found adjacent to the subepithelium, at the base of
the epithelial layer. These basal cells look very
similar to neuron cells in the subepithelium, but
there do not appear to be any connections between
them. The basal cells are strikingly similar to the
basal cells in vertebrate olfactory epithelium
(Graziadei and Monti Graziadei, 1978; Monti Graziadei
and Graziadei, 1979). These cells occur most frequently
in the GRR, are less frequent in the PR and MR, and even
less frequent in the UR. This is indirect evidence of
a receptor or receptor-related function. In the last
three regions the basal cell is noted to occur at or
near a receptor. (In the GRR both basal cells and
receptors are everywhere.) In terms of the similarity
between Lymnaea basal cells and vertebrate basal cells
some very interesting hypotheses can be made as to
their role. Monti Graziadei and Graziadei (1979), in
their study on vertebrate olfactory epithelium, suggest
that the basal cells are involved in the regeneration
of receptor elements after axotomy. (To date, this is

only example of this type of neuron regeneration.) As
far as the comparison can be made, the basal cells
described by Monti Graziadei are nearly identical to
those found in Lymnaea. The ultrastructural charac-
teristics of the cells and their placement in the
epithelium are the basis for this comparison. One
thing that was noted during the work with Lymnaea was
the ease with which the epithelium could be destroyed
by accidental touching. This lead to the thought that
even though there is a protective mucous covering on
the snail, it is possible that scraping the mouth area
on a variety of surfaces, which is a frequent activity
of the snail, could at times be destructive. Consid-
ering this and normal cell renewal, it could be
hypothesized that Lymnaea's basal cells have a role in
renewal, the same function of those in vertebrate
systems. Several casual observations, including
placement of basal cells, their physical relationship
to ganglia, neuron endings, and their nearly identical
structural characteristics to the cells in the ganglia,
contribute to this thought. Pursuit of this hypothesis,
though, was beyond the scope of this study.
One other possible hypothesis for basal cells is
an intraepithelial receptor. Again, similarities to
the neuron cell bodies suggest this hypothesis, but no
other evidence, such as endings extending to the

surface, was found to support it during the course of
this study.
Curly Microvilli
The most interesting feature in the GRR is the
presence of curly microvilli (CMV). These are micro-
villi which are thinner than their straight counter-
part (SMV), and are bent and twisted around each other
to form a sort of "spongy" layer. They are found only
in the GRR and are not found on receptor endings.
Zylstra (1972a) mentions the presence of a spongy
layer above the ganglia of the tentacle and that they
occur on the tips of tentacles of other snails and
slugs. They are not mentioned in the description of
the microvilli of the lip. Wright (1974) describes
a similar layer for the slug Arion ater. Zylstra
(1972a) only suggests they may play a role in sensory
function. A more in-depth speculation as to their
function may be made based on consideration of several
points: 1) They are only found in the GRR; 2) they
are found primarily on support cells; 3) they are found
in conjunction with, but not usually on, sensory
receptors. These points suggest that the CMV may be
involved in trapping molecules in the mucous layer for
concentrating and easier perception by the receptor
endings. It is possible that they provide a concentra-

ting mechanism, more effective than that of the mucous
layer, perhaps making the specialized receptors more
efficient. Or, it is possible that they play a more
passive, protective role.
Distribution of Cells
Although the frequency for each cell varies,
in general, the distribution of El, E4, and basal cells
is as follows: most frequent in the GRR, less in the
PR, even less in the MR, and rarely, if at all, in
the UE. E2, E3, E5, and E6 follow a similar pattern,
but are not so well localized. These frequencies are
the same patterns which are shown by ganglia and
disjoined neuron cells combined. This provides
indirect evidence that the receptor is a process from
a neuron cell body located in the subepithelium.
The greatest number of receptor endings, in
both density and type, are indeed found in the GRR.
This region then is hypothesized to be the area of
the lip most involved in perception of the environment.
While it is not surprising, it is none-the-less quite
interesting to note that the epithelium directly above
the ganglia consists almost entirely of receptor
endings and appears to have few support cells. Table 4
summarizes the types of cells discussed in this paper.
In the UE, only three cells discussed here are