J. Cell Sci. 61, 273-287 (1983)
Printed in Great Britain © The Company of Biologists Limited 1983
273
RELATIVE DAUGHTER CELL VOLUME AND
POSITION OF THE DIVISION FURROW IN
TETRAHYMENA
K. K. HJELM
Department of Molecular Biology, Odense University, Campusvej 55,
DK-5230 Odense M, Denmark
SUMMARY
The relative daughter cell volume (RDCV) values for Tetrahymena pyriformis were determined
at division on live cells. It was found that the anterior cell is generally larger than the posterior cell,
and that the RDCV values are distributed in groups 5-6% apart. The RDCV value was found to
be independent of predivision cell volume, indicating that the mother cell is divided into proportional volumes. The cells seem, however, not to assess volume directly but rather a parameter related
to the cell volume. Furthermore, the RDCV value was found to increase during cell division, so that
the final value is not reached until actual separation of daughter cells.
It is suggested that the division furrow is positioned so that the area of the cell surface lying
between the old oral apparatus and the posterior pole of the cell is divided into equal parts. It is
further suggested that several alternative values of the RDCV are possible, only one of which is
expressed in each cell. The early division furrow is placed anteriorly to its final position, and its
location is adjusted during cytokinesis.
INTRODUCTION
The assembly of cellular structures and organelles from their component molecules
is a late step in the expression of the genetic information of cells. For some organelles
like ribosomes or mitochondria this information determines only their assembly
precisely, while their number may vary within wide limits. For other structures,
however, the cell contains additional inherited information determining precisely
their number, position and polarity. A good example is the oral apparatus of ciliates.
This is a single structure with a definite position and a clear polarity relative to the rest
of the cell. A further point of interest is that in the ciliates, at least, this inherited
information on number, position and polarity may not be contained in the base
sequence of nucleic acids. Several experiments indicate that it is connected instead
with pre-existing structures in the cell (Nanney, 1968; Sonneborn, 1970).
Very little is known about the mechanisms involved in positioning of cellular
structures, but the ciliates, which are especially suited for such investigations, are
being intensively studied to obtain fundamental data on these problems (see reviews
by Aufderheide, Frankel & Williams, 1980; Nanney, 1980). The purpose of the
present work was to obtain basic information on the position of the division furrow
in Tetrahymena, as measured by the value of the relative daughter cell volumes. The
furrow was chosen because it is a single structure, placed in a definite position, and
common to most cells.
274
MATERIALS
K. K. Hjelm
AND
METHODS
Organism
Tetrahymena pyrifonnis from exponentially multiplying cultures was used in all experiments. It
was grown at 28°C in the proteose-peptone, yeast extract medium described by Hjelm (1970),
except that the amount of KH2PO4 was reduced to 25 %. The average generation time was 2-5 h.
Synchrony of division was induced by exposing rotating cultures to heat shocks as described by
Hjelm (1970).
Cells used for volume determination were gently isolated with a micropipette and maintained at
28 °C in medium from their original culture.
Determination of cell volume
The volume of free-swimming cells was calculated from cell length and width measured on
photomicrographs by using the formula for a prolate spheroid. In all other experiments live cells
were flattened between two glass slides, and photomicrographed using a microflash through a 40X
objective (10X was used in Figs 1-3). Three enlarged prints with a total magnification of 1000X
(a few hundred times were used in Figs 1-3) were cut out and weighed. The value of the relative
daughter cell volume was calculated as the ratio of the average weight of the prints of the proter
(anterior daughter cell) to that of the opisthe (posterior daughter cell).
Two methods of flattening the cells were used. In initial experiments (Figs 1-3) a piece of glass
cut from a microscope slide and supported by Vaseline was pressed down on the cells. In later
experiments cells were transferred in a negligible amount of culture medium into a known volume
of liquid paraffin placed on a microscope slide. A coverslip was then lowered onto the oil and the
cells were gradually flattened to a thickness (nominally 7-lO^im) determined mainly by the volume
of the oil and the area of the coverslip. The true thickness is likely to be lower because some of the
oil flows outside the edge of the coverslip. This 'oil method' allowed quick photomicrography of
cells, because they were 'automatically' flattened, it minimized errors due to non-parallel glass
surfaces, and allowed a rough estimate of the maximum absolute volume of a cell by multiplying its
area by its nominal thickness. The area was calculated from the weight of the print and the paper
weight per unit area.
Due to the close contact between cells and the glass surfaces these had to be scrupulously clean.
A generation time of 2-5—3-5 h for the flattened cells was taken to indicate sufficient purity.
RESULTS
Comparison of methods
To avoid possible artifacts due to fixation it seemed preferable to determine cell
volumes on live Tetrahymena. This may be accomplished by measuring the cell axes
and using the formula for a prolate spheroid (e.g., see Scherbaum, 1956). Alternatively, Tetrahymena may be flattened between two glass slides and the volume calculated
from area and thickness (e.g., see James & Reed, 1957; Cameron & Prescott, 1961;
Summers, 1963).
To compare the two methods the values of the relative daughter cell volumes
(RDC V) were determined repeatedly on dividing cells, first when free-swimming and
then as they were more and more flattened (Fig. 1). The results show that, except for
cell no. 3, measurements on the free-swimming cells produced the greatest variation.
The photomicrographs indicate that this is caused by their deviation from a prolate
spheroid. All further measurements were, therefore, made on flattened cells. In Fig.
1 the maximum difference between successive RDCV values of the same flattened cell
is below 2% in six of the cells and does not exceed 4-2% in any cell.
Relative daughter cell volume in Tetrahymena
275
1-36
1
2
3
4 5 6 7 8
Cell number
9
10
Fig. 1. Relative daughter cell volume values determined for 10 dividing cells. Each cell
was photographed repeatedly within a few minutes, first when free-swimming (D), and
then as it was more and more flattened ( • ) . The areas of the pictures of the flattened cells
were 1-1-1-8 times that of the free-swimming cells. Among the photographs of each freeswimming cell only those showing the cell in different stages of rotation around its longer
axis were used.
Relative volume of daughter cells
This was measured on a number of cells with the results shown in Fig. 2. In more
than 90 % of the cells the proter is the largest. The RDCV values range from 0-96 to
1-14 or even up to 1-25, corresponding to a maximum difference of 17-30%.
If the two daughter cells were unequally flattened due to non-parallel slides, or if
they were of different mechanical rigidity, cytoplasm might have been squeezed from
one daughter cell to the other. To exclude this, some of the cells in Fig. 2 were
flattened and photographed immediately after completed division. To be able to
identify the proter and opisthe, the mother cells were incubated with india-ink and
photographed immediately before division. The RDCV value was included in Fig.
2 only if the proter and opisthe contained clearly different numbers of food vacuoles.
The resulting 24 cells show the same general distribution as the rest of the cells.
276
K. K. Hjelm
10
"55
u
° 6
-Q
E
i 4
0-96
0-98
100
102
104
106
108
1-10
1-12
Relative daughter cell volume (proter/opisthe)
M 4 V 1-24
Fig. 2. Distribution of the RDCV values of 84 cells in late or completed division. RDCV
values offlattenedcells from Fig. 1 are included. Broken lines indicate the number of cells
within each class that had their RDCV value determined immediately after completed
division.
Furthermore, the RDCV values obtained on free-swimming cells (Fig. 1) and other
cells (not shown) support the conclusion that in Tetrahymena the proter is in general
larger than the opisthe.
Absolute or proportional assessment of the RDCV values?
In order to make models of the positioning of structures like the division furrow it
is essential to know if it is placed at a certain absolute distance from a given reference
point, regardless of the size of the cell (absolute assessment), or if it is placed so that
it divides the cell into proportional parts regardless of cell volume (proportional
assessment).
To test this the RDCV value was determined on cells synchronized for division.
The mean volume of these cells is two to three times that of exponentially growing
cells (Scherbaum, 1956), and a clear change in RDCV value would be expected if it
is influenced by predivision volume. However, the results (Fig. 3) are similar to those
of Fig. 2. The proter is largest in 80% of the cells, and only approximately 10% of
the cells have an RDCV value clearly different from (i.e. larger than) the values shown
in Fig. 2. Thus the results in Fig. 3 indicate that proportional assessment is used to
determine the RDCV value and the position of the division furrow.
To see if this could be confirmed in exponentially multiplying cells, their RDCV
value was determined with the more precise oil method. Only newly separated
daughter cells were used, so that final RDCV values were determined. This also
eliminated any doubts about where to cut through the furrow on the prints of the
dividing cells. At this time it had been noticed that the newly formed proter may be
distinguished from the opisthe because it is more pointed at the anterior end and/or
-
Fig. 3. Distribution of the RDCV values of 105 cells in late or completed division. The cells were synchronized for division. Broken lines
indicate the number of cells within each class that had their RDCV value determined immediately after completed division.
Relative daughter cell volume (proter/opisthe)
278
K. K. Hjelm
has the contractile vacuole closer to the posterior end. Feeding with india-ink,
therefore, was not necessary. The results (Fig. 4) support the conclusion that
Tetrahymena uses proportional assessment to determine the RDCV value. The
volume of predivision cells varied by a factor of approximately 2 (Fig. 5). If the
RDCV value was determined by absolute assessment it would differ by a factor of
at least 3 and this is clearly not the case (Fig. 4). Besides, Fig. 4 shows, like Fig.
2, that the proter is in general larger than the opisthe, although this applies to only
65 % of the cells. The variation in RDCV values is similar to that shown in
Fig. 2.
Although absolute assessment of the RDCV value is excluded, the cells do not
necessarily divide in strictly proportional volumes. A small fraction of the cell might
not be included in the proportionally divided part. If the volume of this fraction does
not increase in proportion to that of the rest of the cell, the RDCV value will not be
completely independent of the predivision volume. To test this the absolute volumes
of the cells shown in Fig. 4 were plotted against their RDCV value (Fig. 5). Linear
regression analysis of the cells having RDCV values below and above 1-00, respectively, showed no correlation with the absolute volume, and no non-linear relationship
is indicated visually. Thus, within the accuracy of the method the RDCV value seems
completely independent of the predivision cell volume.
Most of the volumes represented in Fig. 5 are between 15 000 and 30 000jUm3. A
variation of approximately four times (from 7500 to 30 000 jtim ) would then be expected in a population containing pre- and postdivision cells. This agrees well with the
findings of Ricketts & Rappitt (1974), who used electronic volume determination, and
Scherbaum (1956), who measured single cells, and indicates that the cells used in the
present study are representative of the whole population.
10
= 8
O
E
0-90"
0-96 0-98
100
102 104
106 108
Relative daughter cell volume (proter/opisthe)
MO
1-12
Fig. 4. Distribution of RDCV values determined on 68 cells immediately after completed
division.
Relative daughter cell volume in Tetrahymena
0-94
0-96
0-98
100
102
104
106
108
Relative daughter cell volume (proter/opisthe)
1-10
279
1-12
Fig. 5. The relationship between the RDCV values and the sum of the absolute volumes
of the two daughter cells determined immediately after completed division. Cells are
identical to those of Fig. 4.
Discontinuous variation in RDCV value
The results of Figs 2-4 indicate that the RDCV value varies in steps rather than continuously. Fig. 2 indicates maxima at 1-05 and 1-095. In Fig. 3 maxima at 0-975, 1-025
and 1 -09 are more evident. In both figures changes in class width or limits result in less
clear maxima, indicating that they may not be significant. Therefore, a further reason
for doing the experiments depicted in Fig. 4 was to see if the discontinuous variation
could be confirmed by these more precise measurements. They resulted in clear maxima at 0-985 and 1-045, which persist, regardless of changes in class-width and limits.
280
K. K. Hjelm
1-10 -
0-88 -
-20
-10
0
+10 +20
Time before and after
separation of daughter cells (min)
Fig. 6
Figs 6-8. Changes in the RDCV value during division of individual cells (each shown by
a different symbol). Cells were flattened either shortly before (Fig. 6) or shortly after (Fig.
7) early furrow formation, or shortly after the previous division (Fig. 8). Fig. 8 is divided
in two parts to facilitate display. Omin is the time of separation of daughter cells.
The distance between the maxima in Fig. 4 is 6% and, considering the accuracy
of the measurements, this agrees rather well with the distance of 4-5 % between the
maxima in Fig. 2 and the distance of 5 % and 6-5 % of the maxima in Fig. 3. Furthermore, a maximum at approximately 0-98 is found in both Figs 3 and 4, another is seen
at approximately 1-045 in Figs 2 and 4, while a maximum at approximately 1-095 is
common to Figs 2 and 3. The maximum at 1-025 in Fig. 3 is the only one not repeated
in another figure. This agreement in separation and position of most maxima further
supports the idea that they are real. Taken together, Figs 2-4 indicate that the
Relative daughter cell volume in Tetrahymena
281
RDCV value and thus the position of the division furrow does not vary continuously,
but rather in steps separated by 5-6% of the RDCV value.
The results in Fig. 4 originated from 10 different cultures. For two of these the
RDCV values fell predominantly in different maxima, and this pointed to the possibility that the RDCV value might depend on some uncontrolled culture condition.
However, for the rest of the cultures no correlation was found with the RDCV values.
When is the final RDCV value determined?
Hypothetically, the RDCV value may be determined either when the division
furrow is initiated or when-the daughter cells separate. To obtain information on this
point the RDCV value was followed during division. Initial experiments showed that
flattening at the earliest step of furrow formation often caused the furrow to regress.
Therefore, cells were flattened either before or after this period. The results show that
the RDCV value tends to increase during division (Figs 6-7). The irregular shapes
of the curves may reflect the fact that cells were somewhat disturbed by the flattening
around the time of constriction.
-10
0
+10 +20
Time before and after
separation of daughter cells (min)
-20
Fig. 7. For legend see p. 280.
CEL61
282
K. K. Hjelm
In another experiment the RDCV value was followed during division of cells that
had remained flattened since their previous division. These cells grow only a little
slower than cells in exponentially multiplying cultures and seem to divide normally.
The results clearly show that the RDCV value increases as division proceeds (Fig. 8).
Taken together, Figs 6-8 strongly indicate that the final RDCV value is not reached
until the daughter cells separate. This demonstrates the importance of measuring the
RDCV value at a well-defined time during division. The figures also indicate that the
increase in RDCV value during division varies from cell to cell, and that the division
furrow is placed initially in a position anterior to that corresponding to the final
RDCV value.
Volume, area or distance assessment?
It is not known if the RDCV value is determined by direct assessment of volume,
1-12 -
0-92 - 1 5 - 1 0 - 5 0 +5
-15-10-5
+5
Time before and after
separation of daughter cells (min)
Fig. 8. For legend see p. 280.
Fig. 9. Distribution of the RDCV values determined on 83 cells immediately after completed division. The cells had
been kept flattened since the preceding division. Broken lines indicate the number of cells within each class that had
a generation time of 2.5 h.
Relative daughter cell volume (proter/opisthe)
284
K. K. Hjelm
or if it is determined indirectly by assessing, for example, the surface area of the cell,
or the distances from either the poles or from some other reference points.
If the cells assess volume directly, the RDCV value would be independent of the
shapes of the anterior and posterior parts of the predivision cells, while the opposite
is true if relative area of distance is assessed. This was used in an experiment where
cells were kept flattened from one division and throughout the next, when their
RDCV value was determined immediately after separation. These flattened cells are
more rounded in the anterior end than the free-swimming cells but seem to grow and
divide normally, although often with a prolonged generation time. Only cells that
divided within 3-5 h of their previous division were used. The results show that most
RDCV values fall between 1-02 and 1-24 (Fig. 9). These are clearly higher than the
RDCV values shown in Figs 2—4. This indicates that the cells do not measure volume
directly, provided, of course, that the measuring mechanism itself is not disturbed by
the flattening. Maxima like those of Figs 2—4 are not evident in Fig. 9.
DISCUSSION
The results of Figs 2-4 and 9 all show that proters are in general larger than
opisthes. Is this due to a systematic error of measurement? The anterior end of the
proter is often more pointed and seems more rigid than the anterior end of the opisthe
and might, therefore, not be in contact with the glass when flattened. This would
result in a seemingly larger proter. However, since the volume anterior to the oral
apparatus amounts to only a small percentage of the total cell volume, the RDCV
value would hardly change by more than 1—2%.
The amount of macronuclear DNA in daughter cells shows a mean difference of
8-10% (Cleffmann, 1968; Doerder, Frankel, Jenkins & DeBault, 1975). This is
slightly more than the difference reported here for the RDCV values (Figs 2, 4). So
far, the RDCV value and the DNA content have been determined on the same cells
only in a conical mutant of Tetrahymena. Here they showed no correlation (Doerder
et al. 1975). This is in agreement with the observation that the constriction of
macronucleus and cytoplasm are not necessarily spatially related (Gavin & Frankel,
1966; Doerder et al. 1975).
The separation of the maxima in Figs 2-4 is small relative to the accuracy of the
method (Fig. 1), and more measurements with improved precision would clearly be
desirable. However, their similar distance and position, and their clear separation
after use of the improved oil method (Fig. 4), are strong indications that they are real.
One obvious error can be excluded: namely, that the proter and opisthe of true
maxima were sometimes accidentally interchanged so that a false extra maximum at
the reciprocal RDCV value would arise. The only maxima fitting this possibility are
those of Fig. 3 at 1-025 and 0-975, but here most of the cells were photographed before
separation, so that the proter and opisthe could not have been mistaken.
If the maxima are real, they have some interesting consequences. A shift in RDCV
value from 0-985 to 1-045 (Fig. 4) requires that the plane of division is moved only
1 %, or 0-6jUm, along the length of a medium-sized predivision cell (assumed to be
Relative daughter cell volume in Tetrahymena
3
285
a prolate spheroid of 18000jUm with a ratio of length to width of 2-6). As a
consequence, it seems unlikely that the final RDCV value is determined when the
division furrow is initiated (for example, because it is simply initiated before or after
a certain ciliary unit). The furrow would have to be placed with an accuracy of a
fraction of the 0-6/im, i.e. 0-1-0-2^m, and constriction of the cell would have to
proceed with similar precision. Otherwise, a clear separation of the maxima could not
be expected. That the furrow should be able to move with this precision seems even
less likely in view of the thickness (>0-5/im) of the contractile ring of the furrow
region (Yasuda, Numata & Watanabe, 1980; Jerka-Dziadosz, 1981). Alternatively,
one could suggest that the furrow is placed initially in an approximately correct
position and that it is adjusted as needed during constriction, so that the 'desired' final
RDCV value is obtained at separation. The results of Figs 6—8 are in agreement with
this reasoning. They show that the RDCV values increase during constriction, in a
different way in different cells, and that the final RDCV value is not established until
the daughter cells separate.
The results of Figs 3-5 support the view that the RDCV value is determined by
proportional assessment. This conclusion is similar to that reached by Lynn & Tucker
(1976) for the positioning of the new oral apparatus in Tetrahymena.
A separate problem is which parameter (volume, area or distance) the cell determines when assessing the RDCV value. Measurement of relative distances from the
poles to the furrow seems less likely, because variations in the relative shape of the two
halves of the cell would result in relatively large changes in the RDCV value. This
would hardly allow the closely spaced maxima of Figs 2—4. Provided that prolonged
flattening of the cells does not induce changes in the mechanism determining the
RDCV value, the results of Fig. 9 indicate that volume is not measured directly. This
leaves assessment of the surface area as the most obvious alternative. This agrees with
the proposal of Lynn (1977), that assessment of the surface area of the cell is used in
determining the distance between the old and new oral apparatus. It would seem
'economical' for the cell to use the same mechanism to position both the new oral
apparatus and the division furrow, as the two always appear close to one another.
The most likely reference point for the positioning of the new oral apparatus is the
pre-existing apparatus (Lynn, 1977). If this is assumed to be the reference point also
for determining the RDCV value, the volume of the cell anterior to the oral apparatus
may not be taken into account when the cell divides. This volume is a small percentage
of the total cellular volume, and if the rest of the cell is divided equally, an RDCV
value slightly larger than unity would be expected. This is in agreement with the
RDCV values found for most cells (Figs 2-4). Furthermore, if the volume anterior
to the oral apparatus increases in proportion to that of the whole cell, the RDCV value
would still be independent of the predivision cell volume, as indicated by Fig. 5.
If it is further assumed that the mechanism determining the RDCV value creates
a potential for several discrete alternative values, and that other factors determine
which of these becomes expressed in each cell, an RDCV value below 1-0 as well as
the maxima of Figs 2—4 can be explained. Unequal division as reported for the conical
mutant (Doerder et al. 1975) may then be due to expression of one of the more
286
K. K. Hjelm
extreme possibilities of the RDCV values. Very unequal divisions, as found in the
monstrous polymorphic Tetrahymena (Hjelm, 1977), may, in addition, involve
displaced reference points.
Summing up, it is suggested that the mechanism positioning the division furrow
creates a potential for several alternative final RDCV values, only one of which
becomes expressed in each cell. The furrow is initiated at a position anterior to that
corresponding to the final RDCV value and is adjusted during constriction of the cell.
The RDCV value is determined by proportional division of the cell surface area
situated between the old oral apparatus and a posterior reference point, close to or at
the posterior pole of the cell.
It is difficult to suggest how the potential for the alternative RDCV values may
arise, especially because very little is known about the mechanisms involved in
positioning of cellular structures in general. However, one possible explanation is that
it arises by interference between waves that are in or out of phase at regular intervals
at the equatorial zone of the cell (Goodwin & Cohen, 1969). It is worth noting that
ciliates possess wave-producing mechanisms at their surface, as is shown by their
ciliary beats.
I should like to thank Mrs I. Hansen for patient and careful technical assistance and Professor Leif
Rasmussen for critically reading the manuscript.
REFERENCES
AUFDERHEIDE, K. J., FRANKEL, J. & WILLIAMS, N. E. (1980). Formation and positioning of
surface-related structures in Protozoa. Microbiol. Rev. 44, 252-302.
CAMERON, I. L. & PRESCOTT, D. M. (1961). Relations between cell growth and cell division. Expl
Cell Res. 23,354-360.
CLEFFMAN, G. (1968). Regulierung der DNS-Menge im Makronucleus von Tetrahymena. Expl
Cell Res. 50, 193-207.
DOERDER, F. P., FRANKEL, J., JENKINS, L. M. & DEBAULT, L. E. (1975). Form and pattern in
ciliated protozoa: Analysis of a genie mutant with altered cell shape in Tetrahymena pyrifonnis,
syngen \.J. exp. Zool. 192, 237-258.
GAVIN, R. H. & FRANKEL, J. (1966). The effects of mercaptoethanol on cellular development in
Tetrahymena pyriformis. J. exp. Zool. 161, 63-82.
GOODWIN, B. C. & COHEN, M. H. (1969). A phase-shift model for the spatial and temporal
organization of developing systems.X theor. Biol. 25, 49-107.
HJELM, K. K. (1970). A technique for cultivation of Tetrahymena in rotating bottles. ExplCellRes.
60, 191-198.
HJELM, K. K. (1977). Monstrous Tetrahymena with intraclonal variation in structure produced by
hereditary modification of normal cells. J . Protozool. 24, 420-425.
JAMES, T. W. & READ, C. P. (1957). The effect of incubation temperature on the cell size of
Tetrahymena pyriformis. Expl Cell Res. 13, 510-516.
JERKA-DZIADOSZ, M. (1981). Cytoskeleton-related structures in Tetrahymena thermophila:
Microfilaments of the apical and division-furrow rings. J. Cell Sci. 51, 241-253.
LYNN, D. H. (1977). Proportional control of organelle position by a mechanism which similarly
monitors cell size of wild type and conical form-mutant Tetrahymena. jf. Embryol. exp. Morph.
42, 261-274.
LYNN, D. H. & TUCKER, J. B. (1976). Cell size and proportional distance assessment during
determination of organelle position in the cortex of the ciliate Tetrahymena. J. Cell Sci. 21,
35-46.
NANNEY, D. L. (1980). Experimental dliatology. New York: Wiley-Interscience.
Relative daughter cell volume in Tetrahymena
287
NANNEY, D. L. (1968). Cortical patterns in cellular morphogenesis. Science, N.Y. 160, 496-502.
RICKETTS, T . R. & RAPPITT, A. F. (1974). Determination of the volume and surface area of
Tetrahymena pyriformis, and their relationship to endocytosis. jf. Protozool. 21, 549-551.
SCHERBAUM, 0 . (1956). Cell growth in normal and synchronously dividing mass culture of
Tetrahymena pyriformis. Expl Cell Res. 11, 464-476.
SONNEBORN, T . M. (1970). Gene action in development. Proc. R. Soc. Lond. B, 176, 347-366.
SUMMERS, L. G. (1963). Variation of cell and nucleus volume of Tetrahymena pyriformis with three
parameters of growth: Age of culture, age of cell, and generation time. J. Protozool. 10, 288-293.
YASUDA, T., NUMATA, 0 . , OHNISHI, K. & WATANABE, Y. (1980). A contractile ring and cortical
changes found in the dividing Tetrahymena pyriformis. Expl Cell Res. 128, 407-417.
{Received 11"June 1982 - Revised 3 November 1982)