A M . ZOOLOCIST, 10:157-165 (1970).
Studies on Regeneration of the Lizard's Tail
SIDNEY B. SIMPSON, JR.
Department of Anatomy, Case Western Reseme University, Medical School.
Cleveland, Ohio 44106
SvNorsis. The present paper is concerned mainly with studies from my laboratory on
regeneration of the lizard's tail. Over the past five years my students and I have
focused our attention on: (1) a careful description of the histology of tail regeneration; (2) the role of neural elements (nerve and ependyma) in the initiation of
regeneration; (3) the role of the ependymal epithelium in the guidance of regenerating central fibeis: and (4) the development of techniques of tissue culture applicable
to this system.
The purpose of this paper is to review
some of the studies from my laboratory on
regeneration of the lizard's tail. I shall
summarize briefly those results that bear
on general problems in vertebrate regeneration. At the same time emphasis will be
placed on areas in which this particular
system may prove advantageous in future
studies. The bibliography is highly selective rather than encyclopedic in nature.
Wherever possible I have cited key papers
or review articles. To those authors who
most certainly deserve citation but were
not acknowledged, the author offers his
apologies. The general features of tail regeneration in the two lizards that we have
studied, Lygosoma laterale (Scincidae) and
An olis carolinensis (Iguanidae), have
been reported (Simpson, 1965; Cox,
1969r/) and will not be restated here.
ROLE OF THE NERVOUS SYSTEM"
The importance of the nervous sytcm in
vertebrate regeneration is, by now, legion
among students of regeneration. With the
exception of the provocative experiments
on aneurogenic Ambystoma limbs, initiated
by Yntema (1959) and extended by Steen
and Thornton (1963) and Thornton (this
Symposium), the regeneration of limbs
and tails in vertebrates is dependent on the
Various phases of the author's work reported
herein were supported by Research Grant GM
12653 from the Division of General Medical
Sciences, USPHS. The author holds a PHS Research
Career Development Award CMI7, 102-02.
presence of an adequate nerve supply (see
reviews of Singer, 1952 and 1965).
In amphibian limb regeneration, the
relationship has been shown to be a quantitative one. Singer (1947), demonstrated
that a given threshold number of nerve
fibers (expressed as number of fibers / (100
n)- cross-sectional tissue area) was necessary for regeneration of the Triturus forelimb. Singer, et al. (1967) have suggested
that the relationship may be expressed
more accurately in terms of total neuronal
volume per (100 n)2. The influence of
nerves would appear to be shared equally
by axon and nerve cell body (Kamrin and
Singer, 1959).
Likewise, the role of the nervous system
in amphibian tail regeneration has been
documented (Goldfarb, 1909; Holtzer,
1959; Kiortsis and Droin, 1961). In these
cases, however, the precise role of the central and peripheral nervous systems has
not been elucidated. The results of these
studies suggested that the spinal cord was
requisite for tail regeneration.
The regeneration of the lizard's tail has
also been shown to be dependent on the
presence of the spinal cord (Kamrin and
Singer, 1955; Simpson, 1962). In both
cases, failure of lizard tails to regenerate
after removing the spinal cord was interpreted to be due to a massive reduction of
the nerve-fiber content of the tail. Experiments of this type only demonstrated that
some portion of the spinal cord was essential.
157
158
SIDNEY B. SIMPSON, JR.
Work with the brown skink, Lygosoma
latemle, (Simpson 1964) suggested that the
nerve-fiber threshold relationship probably did not hold for regeneration of the
lizard's tail. Instead, it was suggested that
the ependymal lining of the central canal
of the spinal cord was responsible for initiating regeneration. Hughes and New
(1959) had even earlier attached some
importance to the very close association between the outgrowing ependymal tube and
the wound epithelium during tail regeneration in Spluierodactyhis. Simpson (1964)
was able to test the importance of
ependymal epithelium vs. nerve. The experiment consisted of taking autografts of
cartilage tube and ependyma from a regenerate and transplanting them into wounds
prepared in the dorsum of the tail. This
procedure takes advantage of the fact that
the regenerated spinal cord of the lizard is
a heteromorphic structure consisting of
ependymal epithelium and descending
central fibers. No nerve cell bodies are
present. Therefore, removing a segment of
the regenerated cord results in section of
the descending central fibers and their degeneration within a period of four to five
days. Grafts have been made in which the
ependyma, along with the cartilage tube,
were placed in dermal pockets for five days
(allowing degeneration of nerve fibers)
and then removed and placed in dorsal
wounds in the tail. This procedure removes the possibility of some residual influence of the sectioned nerve fibers. The
dorsal wounds were prepared in such a
way that the spinal cord of the tail in this
region was not injured. The results obtained were very clear-cut. Autografts of
cartilage tube and ependyma resulted in a
complete regenerate from the dorsum of
the tail. Control grafts of cartilage tube
alone never resulted in regenerates.
These experiments most certainly demonstrated that: (1) the nerve-fiber contribution
from the spinal cord is not necessary for regeneration; and (2) that the ependyma
was probably the important factor. However, assessment of the possible impor-
tance of peripheral nerve fibers was
not made in these experiments. Recently,
Cox (1969b) has investigated this problem
in the lizard, Anolis. He found that the
total number of peripheral nerve fibers per
(100 p)- surface area in the normal tail
was between 4.1 and 4.5. Removing dorsal
root ganglia for three segments anterior to
a given plane of autotomy reduced the
peripheral count to an average of 0.31 to
0.34 fibers/ (100 M ) 2 . This, however, did not
inhibit regeneration. Ablation of the
spinal cord for three segments anterior to
the plane of autotomy only reduced the
peripheral count to 2.8 fibers/ (100 ^) 2
and regeneration did not occur. Thus, reduction of the peripheral count to less
that one fiber/ (100 jj)- does not alter regeneration. This, coupled with the fact
that ependymal grafts in Anolis (without
contribution from central fibers) can induce regenerates, clearly substantiates the
role of the ependyma. This relationship
may well hold for amphibian tail regeneration as well. Kiortsis and Droin (1961)
have shown that spinal ganglia or brain
tissue can not be substituted for spinal
cord in regeneration of the tail of the European newt.
The mechanism of the ependymal influence is difficult to explain. Singer, (1952,
1905) has suggested that the influence of
nerves in regeneration is clue to the production of a 'trophic substance' by the cell
body and its transmission via the axon to
the tissues involved. This 'trophic substance' may be able to pass directly from
the nerve into the surrounding tissue
(Knmrin and Singer, 1959). Since the
ependymal epithelium represents (like
nerve and glia) one of the end products of
the embryonic neuroepithelium, it is not
unreasonable to assume that the ependyma
may be acting in the same way as nerves in
other regenerating systems.
In this connection, grafts of ependyma
have been made to amputated, normally
non-regenerating limbs in the lizard, Lygosoma. Such implants resulted in the production of a blastema that expressed lim-
REGENERATION OF LIZARD'S TAIL
159
IB
FIG. 1. Induction o£ a regenerative response in the
normally non-regenerating limb of Lygosoma.
A. Experimental limb with heteromorphic regenerate induced by an autograft o£ cartilage tube and
ependyma. B. Control limb with an autograft ot
cartilage tube alone. Lines indicate level o[ amputation.
ited growth (Fig. 1A). Control implants of
cartilage tube did not stimulate regeneration (Fig. IB). These ependyma-induced
outgrowths are as good if not better than
the regenerates induced by nerve augmentation. There is unfortunately little chance
for normal morphogenesis in the ependyma-induced regenerates. This is due to
the fact that a cartilage tube always differentiates iix association with the outgrowing
ependyma.
Making the assumption that the ependyma operates in the same "trophic" way
as the nerves, then the studies on the lizard
suggest that the hypothesized trophic substance may not be unique to cells specialized for conduction of impulses. This has
been suggested earlier by Singer (1965).
phic and deficient in many respects. Although there was a consensus that an
ependymal tube and a portion of the meninges regenerated, there was considerable
divergence of opinion concerning the regeneration of descending central nerve
fibers. Hooker (1912) described regenerated central fibers, and their presence was
later confirmed using selective nerve stains
(Kamrin and Singer, 1955; Simpson,
1964).
We now have a more critical evaluation
of the content and arrangement of the regenerated cord in the lizard, Anolis (Simpson, 1968; Egar, et al., 1969). From these
studies we now know that the regenerated
cord of Anolis consists of: (1) ependymal
epithelium separated from the surrounding meninges by a basement lamina; and
(2) several thousand regenerated descending central fibers, the majority of which
are unmyelinated (Fig. 2). No nerve cells
are regenerated. Some question still exists
as to whether or not glial cells differentiate
within the regenerate. Egar, et al. (1970)
REGENERATION OF THE SPINAL CORD
Several conflicting descriptions of the regenerated lizard cord have appeared in the
literature (Misuri, 1910; Hooker, 1912;
Stefanelli, 1950). These authors agreed
that the regenerated cord was heteromor-
180
SIDNEY B. SIMPSON, JR.
BL
MV
JC
MV
REGENERATION' OF LIZARD'S T A I L
161
FIG. 2. Diagrammatic representation of a segment
of the regenerated cord, illustrating the interrelationships of the various components. Portions of
four ependymal cells are shown. The basal processes of each ependymal cell project to the basement lamina (BL). Large fascicles of regenerated
descending central nerve fibers (NF) are present
between the basal projections of the ependymal
cells. Small lateral projections of the main basal
process extend into the nerve fascicles. Beyond the
basement lamina (BL) is the connective tissue
complex of the regenerated meninges. This complex consists of collagenous fibers, fibrocytes, and
occasional bundles of regenerated central nerve
fibers surrounded by Schwann cells. Lumen
(L), Schwann cell (SCH), basement lamina (BL),
fibrocyte process (FI), fascicle of nerve fibers
(N.F.), nucleus of ependymal cell (N), micropinocytotic vesicles (MPV), miltivesicular body (MVB),
junctional complex (JC). Golgi zone (G), microvilli (MV), cilium (C).
found an occasional cell type that differed
in morphology from the ependymal cells.
They have suggested that these oells are
astrocytes.
Contrary to the observations of earlier
workers (Stefanelli, 1950) the regenerated
central fibers of the regenerated cord do
not undergo atrophy. Regenerates, known
to be six months old, do not exhibit a decreased number of regenerated fibers. This
particular situation poses some interesting
questions for the neurobiologists. Why do
several thousand central fibers regenerate
in a situation where they make no connections with peripheral tissues? What is their
function? Why are they maintained, once
regenerated?
This particular system also offers experimental advantages to workers interested in
the ultrastructure and physiology of nerve
growth cones. Just behind the advancing
ependymal tip, one has the advancing tips
(growth cones) of several thousand fibers.
All of these fibers are oriented longitudinally, and the picture is uncomplicated by
glia and extraneous nerve fibers.
One of the most interesting observations
to come out of these studies centers on the
relationship between the ependymal cells
and the descending central nerve fibers
that have regenerated from the old cord.
During regeneration of the cord, the ependymal cells close over the severed end of
the central canal and form a closed tube
with a slightly dilated vesicle at the distal
end. As regeneration proceeds, the ependymal tube extends into the forming tail
regenerate. The ependymal cells send out
long processes that overlap at the connective tissue interface. Although the
ependymal cells are united by junctional
complexes at their luminal surface, large
spaces are present between adjacent ependymal cell processes. It is into these
spaces that the descending central fibers
regenerate, (Egar, et al., 1970). Thus, in the
lizard, the regenerating ependymal tube
serves as an effective guide for the regenerating central fibers.
Tt has been suggested that the same
mechanism may well acount for the reestablishment of functional connections
during regeneration of ablated portions of
urodelan spinal cords (Simpson, 1968).
Preliminary studies from our laboratory
have in fact shown that this same mechanism of ependymal fasciculation of regenerating central fibers occurs during regeneration of the cord following amputation of
the tail of Triturus.
RESULTS OF POST-URANCHIAL ABLATION OF THE
CORD
Previous studies have shown that regeneration of the spinal cord and return of
function occurs in fish, salamanders, and
larval anurans. Regeneration and functional recovery does not occur in birds and
mammals (see review of Clemente, 1964).
Since the descending central fibers of
Anolis are regenerated during replacement
of the tail, experiments were performed
to determine if regeneration and return of
function would occur following postbranchial ablation. These experiments are
in progress at the present time and the
results to date are preliminary. In five
cases where a 2.0 mm segment of the
spinal cord was removed at the level of the
lower thoracic vertebrae, no functional
162
SIDNEY B. SIMPSON, JR.
perimysiuni? There are also indications in
the literature that this may be the case
(review of Betz, et al., 1966). Indeed, the
muscle of the regenerate may be derived
from a number of sources within the amputated stump.
The relationship between the stump
muscles and the muscle of the regenerate
can be studied to some advantage in the
Lygosorna tail regenerate. Due to the preformed planes of autotomy, separation of
the lizard tail by autotomy does not injure
the remaining stump muscles. Most of the
connective tissue septum that forms the
plane of autotomy remains attached to the
stump muscles. During regeneration,
there is some dissolution of the collagenous
septum, and some of the fibrocytic cells
within the fibrous matrix incorporate tritiated thymidine. However, the septum never completely breaks down and therefore
provides a continuous line of demarcation
between stump muscle and regenerate
proper.
The stump muscles do not undergo any
REGENERATION OF AXIAL MUSCLES
of the classical alterations normally associThe question concerning the origin of ated with morphological (^differentiation
cells that constitute the regenerate is large- of multinucleate striated muscle. None of
ly unanswered. This is true whether one is the muscle cell nuclei has ever been seen
talking about regeneration of limb in a to incorporate labeled thymidine. Thus,
salamander or regeneration of tail in the following autotomy in Lygosonm, the
lizard. This is especially true in the case of stump muscles maintain their integrity
muscle, since muscle of the stump is a during the entire process of regeneration.
complex tissue consisting of muscle fibers, These observations strongly suggest that
connective tissue, and the so called 'satel- there is no direct contribution of regenerlite-cells' closely associated with the muscle ative cells from mature skeletal muscle
fibers. From studies on urodelan limb re- cells within the stump. From where does
generation (see reviews of Hay, 1966, and the muscle of the regenerate arise? The
Goss, 1969) we know that the muscle of situation in the regenerated tail of the lithe stump undergoes changes leading to a zard certainly poses serious difficulties for
loss of normal morphology. Do the mul- the proponents of tissue-specific regenertinucleate muscle fibers release viable cells ation.
The same situation probably holds for
during this process that can go on to proliferate and form the muscle of the regener- regeneration of the tail in the lizard, Anoate? There are numerous precedents in the Hs (Cox, 1969r/). Here, the arrangement of
literature that indicate that this may be so the myotomes is somewhat different from
(Thornton, 1938; Hay, 1962; Steen, 1968). that in Lygosoma. In Anolis the muscle
Does the muscle of the regenerate arise slips are not injured by autotomy; howevfrom multiplication of 'satellite-cells' or er, several muscle slips project past the
connective tissue cells of the endo- and epidermis and are included in the clot that
recovery occurred during the four months
that the animals were maintained. At four
months post ablation, every animal exhibited a histological picture typical of
that described for the non-regenerating
mammalian cord. There had been little
proliferation of the ependymal epithelium.
The central canal had not established
continuity, and a fibrocellular scar was interposed between the cut ends of the cord.
Studies of Bodian-stained sections indicated that no major central nerve-fiber
bundles crossed the ablation-gap.
For studies of regeneration of spinal
cord per sc, the lizard system offers an
unique opportunity. In the same animal:
(1) in one case (during tail regeneration)
the ependyma proliferates and guides the
regenerating central fibers; and (2) in another case (post-branchial levels) these
processes do not occur. One would like to
know what factors allow for regeneration
of central fibers in one situation and are
apparently absent in the other.
REGENERATION OF LIZARD'S TAIL
forms. The tips of a few of the muscle
slips are therefore secondarily ablated.
Cox, (I969fl) found that although a small
number of muscle fibers undergo morphological ^differentiation, this is followed
rapidly by repair of the injured segment.
New myotubes formed and the connective
tissue septum was reconstituted over the
injured tip of the muscle slip. The repair
was, histologically, very similar to that seen
in: (1) muscle regeneration following
crush injuries in mammals; and (2) regeneration following mincing in rats, (see
Carlson, this Symposium). Thus, in Anolis
the injured muscle is repaired and a septum reconstituted long before there is any
differentiation of pro-muscle aggregates
within the regenerate proper. When the
muscle aggregates of the regenerate do
form, they do so at some distance from the
reconstituted muscle septum. This particular system allows one to study in the same
animal, two separate modes of muscle regeneration that are spatially and temporally distinct.
Ts the non-participation of stump muscle
unique to the lizard? Several pieces of
information suggest that this is not so.
First, if one amputates (rather than autotomizes) the lizard's tail, muscle is injured
and does undergo the classical changes associated with ^differentiation. Secondly,
Wake and Dresner (1968), have demonstrated that certain salamanders have preformed planes of autotomy in their tails.
Mufti, (1968) working in my laboratory
has recently studied regeneration of the
tail following autotomy in one of these
salamanders, Desmognatliis f. fuscus. Thus
far, he has shown that, just as in Lygosoina, when the muscles of the stump are
not damaged they do not undergo ^differentiation during regeneration.
STUDIES IN VITRO
Tissue culture has made possible many
experiments that in the whole embryo or
organism would have been extremely difficult, if not impossible. Paradoxically, this
technique was initiated in embryonic am-
163
phibian material (Harrison, 1907), and
yet adult amphibian cells have been quite
refractory to manipulation in vitro.
The full advantages of in vitro techniques have, to date, been denied to students of vertebrate regeneration. Some
progress in the art of culturing adult amphibian material has been made in recent
years (Wolf, et al, 1960; Wolf and Quimby,
1964). Most of the attempts to culture
components of urodelan limb blastemas
have achieved, at best, maintenance and
limited outgrowth (Lecamp, 1957; Fimian,
1959; Stocum, 1968). A considerable
amount of regeneration has been obtained
in organ-cultured larval tails of Xenopus
(Weber, 1967).
The application of standard techniques
of tissue culture to the components of the
lizard tail regenerate have been more
rewarding (Simpson and Cox, 1967). From
the early tail regenerate one can now culture the wound epithelium, blastemal
cells, promuscle, procartilage, and ependymal epithelium. These components can
be dissociated into single-cell suspensions,
plated, and subcultured as easily as embryonic chick or mammalian cells. Most, if
not all of the cellular types can also be
cloned. Even more encourgaing, in vitro
differentiation of two types of cell (striated
muscle and pigment cells) has been obtained.
The propagation and differentiation in
vitro of the promuscle cells of the early
regenerate has been studied in some detail
(Cox, 1968). Single-cell suspensions of
promuscle cells could be plated at clonal
densities and give rise to myoblastic colonies. Each colony passed through a cycle
of proliferation, followed by fusion of
myoblast to form striated muscle fibers. Of
particular interest, on about day 8 or 9,
certain cells within the colony assumed a
rounded-up morphology. This change was
progressive and by day 10 or 11, some
50-75% of the cells in a given colony had
assumed this morphology. At this time if
the growth medium was changed to a "fusion" medium, massive fusion followed
164
SIDNEY B. SIMPSON, JR.
within 24-48 hours. Cultures allowed to
grow continuously in growth medium
stretch out again onto the culture dish and
fail to fuse. Some of these mononucleate
cells develop cross striations.
Autoradiographic studies with labeled
thymidine indicate that these rounded-up
cells have either temporarily withdrawn
from the cell cycle or have an extremely
long Gl period. Short term (30 min) exposures to tritiated thymidine revealed
that none of the rounded cells incorporated label, whereas the flattened cells within
the colonies labeled heavily, (Cox, 1968).
Additional studies based on microspectrophotometric determinations have also
demonstrated that the rounded cells represent a Gl population, (Cox and
Simpson, 1970). These results have been
interpreted to mean that the rounded-up
cells represent a postmitotic, prefusion
population of cells. These data fit with the
data obtained from studies on myogenesis in the chick (Marchock and Herrman,
1967; Bischoff and Holtzer, 1969). The
lizard system, however, has several experimental advantages: (1) there is a build-up
of morphologically recognizable prefusion
cells, with fusion (when elicited) occurring
over a relatively short period of time; and
(2) fusion can be controlled by using a
selective culture medium.
The reaction of promuscle cells to longterm culture has also been explored. To
date, three established cell lines have been
derived. Cells in one of the lines have been
carried at clonal or near clonal densities
for approximately 400 generations. This
line is predominantly diploid and is still
capable of forming striated muscle. The
other two lines represent deviate cell lines
that have lost their ability to form skeletal
muscle.
For many years, workers in the field of
regeneration have wanted to label clonal
populations of a particular cell type, introduce them back into a regenerating system
and test their differentiative potential.
The lizard system now provides an opportunity to do just this sort of experiment.
REFERENCES
Betz, E. H., H. Firket, and M. Reznik. 1966. Some
aspects of muscle regeneration. Int. Rev. Cytol.
19:203-227.
Bischoff, R., and H. Holtzer. 1969. Mitosis and the
process oC differentiation of myogenic cells in
vitro. J. Cell Biol. 41:188-200.
Clcmente, D. 1964. Regeneration in the vertebrate
central nervous system. Int. Rev. Neurobiol. 6:
258-293
Cox, 1*. C. 1968. In vitro myogenesis of promuscle
cells from the regenerating tail of the lizard
Anolis carolinensis. J. Morphol. 126:1-18.
Cox, P. C. 1969a. Some aspects of tail regeneration
in the lizard, Anolis carolinensis. I. A description
based on histology and autoradiography. J.
Exp. Zool. 171:127-150.
Cox, ]». G. 19696. Some aspects of tail regeneration
in the lizard, Anolis carolinensis. II. The role of
the peripheral nerves. J. Exp. Zool. 171:151-160.
Cox, P. G., and S. B. Simpson, Jr. 1970. A microphotometric study of myogenic lizard cells grown
in vitro. (Submitted to Develop. Biol.)
Egar, M. W., S. B. Simpson, Jr., and M. Singer.
1970. Ultrastructural studies on the regenerating
spinal cord of the lizard, Anolis carolinensis.
(Submitted to J. Morphol.).
Finiian, J. W. 1959. The in vitro cultivation of
amphibian blastema tissue. J. Exp. Zool. 140:
125-138.
Goldfarb, A. J. 1909. The influence of the nervous
system in regeneration. J. Exp. Zool. 7:643-722.
Goss, R. 1969. Principles ot regeneration. Academic
Press, New York.
Harrison, R. G. 1907. Observations on the living
developing nerve fiber. Proc. Soc. Exp. Biol. Med.
4:140-143.
Hay, E. D. 1962. Cytological studies of dedifferenliation and differentiation in regenerating amphibian limbs, p. 177-210. In D. Rudnick, [ed.].,
Regeneration. Ronald Press, New York.
Hay, E. D. 1966. Regeneration. Holt, Rinehart, and
Winston, New York.
Hoh/.er, H. 1959. The development of mesodermal
axial structures in regeneration and embryogenesis, p. 15-33. In C. S. Thornton, [ed.], Regeneration in vertebrates. University of Chicago
Press, Chicago.
Hooker, D. 1912. Die Nerven im regenerierten
Schwanz der Eidechsen. Arch. Mikrosk. 80:
217-222.
Hughes, A., and D. New. 1959. Tail regeneration in
the Geckonid lizard, Sphaerodactylus. J. Embryol.
Exp. Morphol. 7:218-302.
Kamrin, R. P., and M. Singer. 1955. The influence
of the spinal cord in regeneration of the tail of
the lizard Anolis carolinensis. J. Exp. Zool. 128:
611-627.
Kamrin, R. P., and M. Singer. 1959. The growth
influence of spinal ganglia implanted into the
denervated forelimb regenerate of the newt, Trituriis.J. Morphol. 104:415-439.
REGENERATION OF LIZARD'S T A I L
Kiortsis, V., and A. Droin. 1961. La regeneration
caudale des Urodeles (Induction et reactivity du
territoire). J. Embryol. Exp. Morphol. 9:77-96.
Lecamp, M. 1948. Regeneration chez le Triton in
vitro et in vivo. C. R. Acad. Sci. 226:695-696.
Marchok, A., and H. Herrman. 1967. Studies of
muscle development. I. Changes in cell proliferation. Develop. Biol. 15:129-155.
Misuri, A. 1910. Ricerche sulla struttura della coda
normale e rigenerata nella Lacerta muralis.
Merr. Soc. Zool. Ital. 11:103-135.
Mufti, S. 1969. Re-evaluation of salamander tail
regeneration. (Abstr.) Amer. Zoologist 9:613.
Simpson, S. B., Jr. 1962. Regeneration of lizard extremities. M.S. Thesis, Tulane University, New
Orleans, Louisiana.
Simpson, S. B., Jr. 1964. Analysis of tail regeneration in the lizard Lygosoma laterale. I.
Initiation of regeneration and cartilage differentiation: the role of the ependyma. J. Morphol.
114:425-435.
Simpson, S. B., Jr. 1965. Regeneration of the lizard
tail, p. 431-443. In V. Kiortsis and H. A. L.
Trampusch, [ed.], Regeneration in animals and
related problems. North Holland Publ. Co., Amsterdam.
Simpson, S. B , Jr. 1968. Morphology of the regenerated spinal cord in the lizard, Anolis carolinensis. J. Comp. Neurol. 134:193-210.
Simpson, S. B., Jr., and P. G. Cox. 1967. Vertebrate
regeneration system: culture in vitro. Science 157:
1330-1332.
Singer, M. 1947. The nervous system and regeneration of the forelimbs of adult Triturus. VII.
The relation between number of nerve fibers and
surface area of amputation. J. Exp. Zool.
104:251-265.
Singer, M. 1952. The influence of the nerve in
regeneration of the amphibian extremity. Quart.
Rev. Biol. 27:169-200.
Singer, M. 1965. A theory of the trophic nervous
control of amphibian limb regeneration, including a re-evaluation of quantitative nerve require-
165
ments, p. 20-32. In V. Kiortsis and H. A. L.
Trampusch, [ed.], Regeneration in animals and
related problems. North Holland Publ. Co., Amsterdam.
Singer, M., K. Rzehak, and C. Maier. 1967. Relation
between the caliber of the axon and the
trophic activity of nerves in limb regeneration. J.
Exp. Zool. 166:89-98.
Steen, T. 1968. Stability of chondrocyte differentiation and contribution of muscle to cartilage during limb regeneration in the axolotl (Siredon
mexicanum). J. Exp. Zool. 167:49-77.
Steen, T., and C. S. Thornton., 1963. Tissue interaction in amputated aneurogenic limbs of Amblystoma larvae. J. Exp. Zool. 154:207-221.
Slefanelli, A. 1950. Some comments on regeneration
in the central nervous system, p. 210-211. In Paul
Weiss, [ed.], Genetic neurology. Univ. Chicago
Press.
Stocum, D. L. 1968. The urodele limb regeneration blastema: A self-organizing system. I.
Differentiation m vitro. Develop. Biol. 18:
441-456.
Thornton, C. S. 1938. The histogenesis of muscle in
the regenerating forelimb of larval Amblystoma punctatum. J. Morphol. 62:17-47.
Wake, D. B., and I. Dresner. 1967. Functional
morphology and evolution of tail autotomy in
salamanders. J. Morphol. 122:265-306.
Weber, R. 1967. Biochemistry of amphibian metamorphosis, p. 227-301. In R. Weber, [ed], The
biochemistry of animal development, Vol. 2Academic Press.
Wolf, K., M. C. Quimby, E. A. Pyle, and R. P.
Dexter. 1960. Preparation of monolayer cell cultures from tissues of some lower vertebrates.
Science 132:1890-1891.
Wolf, K., and M. C. Quimby. 1964. Amphibian
cell culture: Permanent cell line Crom the bulldog (liana catesbeiana). Science 144:1578-1580.
Vntema, C. L. 1959. Regeneration in sparsely innervated and aneurogenic forelimbs of Amblystoma
larvae. J. Exp. Zool. 140:101-124.