J. Cell Sci. 25, 111-123 (1977) 111

Printed in Great Britain

THE NUCLEOLUS AND MEIOSIS DURING

MICROSPOROGENESIS IN ENDYMION

NON-SCRIPTUS (L.)

B. T. LUCK AND E. G. JORDANBiology Department, Queen Elizabeth College, University of London,Campden Hill Road, London WS 7 AH, England

SUMMARY

Stages of meiosis from the bluebell Endymion non-scriptus (L.) were studied by electronmicroscopy. The segregated components of the nucleolus at meiotic prophase underwentfragmentation and dissolution at pachytene-diplotene. Nucleoli were absent during bothmeiotic divisions and reformed on the nucleolus organizer into a fibrillar mass from scatteredfibrillar components at the dyad and tetrad stages. It is argued that the fibrillar region showscontinuity through nuclear division though undergoing structural transformations in theprocess. Nucleolar reformation occurs on condensed nucleolus organizers. Processing of theribosomal precursors and the resumption of RNA synthesis is discussed in relation to thedispersal of the nucleolus organizer into the fibrillar region of the reformed nucleolus.

INTRODUCTIONNucleolar ultrastructure during the mitotic cycle has been extensively studied by

various investigators (Lafontaine & Chouinard, 1963; Jordan & Godward, 1969;Pickett-Heaps, 1970; Noel, Dewey, Abel & Thompson, 1971; Chouinard, 1966,1971, 1975; Lafontaine, 1958; Lafontaine & Lord, 1974). It is generally agreed thatthe nucleolus suffers a general dissolution during late prophase of mitosis in higherplants; however, there are differing reports on the mode of its dissolution and thesubsequent fate of the different parts (Lafontaine & Lord, 1969; Chouinard, 1971;Moreno Diaz De La Espina, Risueno, Fernandez-Gomez & Tandler, 1976).

The origin of the material which constitutes the reappearing nucleoli is also still amatter of considerable uncertainty. Cytochemical studies of Allium cepa meristematiccells (Stockert, Fernandez-Gomez & Gimenez-Martin, 1970; Gimenez-Martin, DeLa Torre, Fernandez-Gomez & Gonzalez-Fernandez, 1974; Moreno Diaz De LaEspina et al. 1976) support the earlier view that prenucleolar bodies, seen in thenewly formed nucleus, arise from the so called pre-nucleolar material seen around theanaphase chromosomes (McClintock, 1934; Lafontaine, 1958). However, the contri-bution to the new nucleoli of the so called pre-nucleolar fibrillo-granular materialfound around the anaphase chromosomes is questioned by some workers (Swift, 1959;Chouinard, 1966; Lafontaine & Lord, 1974). The extent to which the new nucleolusis dependent on de novo synthesis has been followed in recent experiments employinginhibitors of RNA and protein synthesis (Gimenez-Martin et al. 1974; Semeshin,Sherudilo & Belyaeva, 1975; Risueno, Moreno Diaz De La Espina, Fernandez-Gomez & Gimenez-Martin, 1976).

i i2 B.T. Luck and E. G. Jordan

The synchrony of the meiotic process and the greater duration especially of thephase of nucleolar breakdown in the extended prophase of meiosis, together with thefact that there are 2 phases of reorganization, one after each division, only one of whichleads to a normal interphase, makes it a useful and interesting system for studies on thenucleolar cycle.

As previously reported (Jordan & Luck, 1976), concomitant with nucleolussegregation during meiotic prophase in Endymion non-scrtptus, the nucleolus organizeremerges to the surface of the nucleolus. In this paper we report the various rearrange-ments of the nucleolar zones in the prophase breakdown and the subsequent reorganiz-ation events after each of the 2 meiotic divisions. This work indicates that some of theconstituents of the new nucleolus are synthesized before nucleolar reorganization.Thus the understanding of the nucleolus becomes not only the understanding of thesynthesis of nucleolar materials but also their dispersal and reorganization at nucleardivision.

MATERIALS AND METHODSEndymion non-scriptus (L.) (Bluebell) plants were harvested in mid-January. The cyto-

logical stage of the anthers in a floret was found from an aceto-orcein squash of one anther.The remaining 5 were sliced into 3% distilled glutaraldehyde in 01 M phosphate buffer,pH 6-8 at room temperature. The material was left in glutaraldehyde for 4-5 h, thoroughlyrinsed in buffer, postfixed for 2-3 h at room temperature in 1% osmium tetroxide in the samebuffer, dehydrated through an ethanol-propylene oxide series and embedded in Araldite.Sections were cut using a diamond knife, stained with aqueous lead citrate and post-stainedwith 10% (w/v) uranyl acetate in methanol (Stempack & Ward, 1964), and examined in anAEI EM 6B electron microscope.

Labelling on figures

acclici

fffgifgo1

accessory nucleolichromatinchromosomecytoplasmic imaginationfibrillar regionfibrillar region fragmentgranular regiongranular region fragmentGolgilightly staining region

UinndnenppbpscmescV

lipid dropletmitochondrionnucleoidnuclear envelopenuclear porepre-nucleolar bodypolysynaptonemal complexremnant nuclear envelopesynaptonemal complexvacuole

Figs. 1-3. Electron micrographs showing nucleolus segregation and fragmentation.Fig. 1. Segregated zygotene nucleolus with the nucleolus organizer in an external

position. An accessory nucleolus is on the surface of the nucleolus organizer andcytoplasmic invaginations are present in the nucleus, x 12500.

Fig. 2. Pachytenc-diplotene. High magnification of nucleolus organizing regionshowing fibrils of 5-125 nm in diameter. The accessory nucleoli have a similarfibril size to the fibrillar region: 5-10 nm, and the fibrils of the darkly staining chro-matin are 10-15 n m m diameter, x 32000.

Fig. 3. Pachytene-diplotene. There is a clear separation of the various parts of thenucleolus; the fibrillar region has fragmented. An accessory nucleolus and a remainingsynaptonemal complex are seen in the diffuse chromatin. Few nuclear pores areobserved in the nuclear envelope, x 11000.

The nucleolus during microsporogenesis 113

I I 4 B. T. Luck and E. G. Jordan

The nucleolus during microsporogenesis 115

RESULTS

Nucleolar segregation, separation and dissolution

Zygotene nucleoli displayed the typical 'segregated nucleolus' appearance and alsohad some accessory nucleoli associated with the nucleolus organizer (Fig. 1). Accessorynucleoli had a similar staining intensity to the fibrillar region (Fig. 1), with fibrilsbetween 5 and 10 nm in diameter. The chromatin fibrils were 10-15 n m m diameter(Fig. 2). By diplotene, the various components had separated (Fig. 3). These separateparts of the nucleolus also showed evidence of fragmentation (Figs. 3-6). During thefinal stages of nucleolar dissolution when the fibrillar and granular zones had movedapart from each other and were breaking up there was no indication of the fate of thegranules or fibrils. The small areas of separate fibrillar and granular zones simplybecame smaller until they eventually disappeared. From the compactness of thefibrillar pieces during this process it seems that fibrils must be leaving the surface toreduce the size of the fragments, rather than gradually dispersing en masse by ageneral disaggregation. The granular region, however, although staying in recogniz-able pieces, does seem to disperse by becoming generally looser until the granulescannot be distinguished from the background. At this time there is an increase in thenumber of granules seen in the nucleoplasm. It is difficult to decide whether or notany of the nucleolar materials are contributing to the matrix of the recondensingchromosomes because they are still very dispersed at this time. During nucleolardissolution large areas of the nuclear envelope remain intact (Figs. 3, 4). Frequently,large pieces of nucleolar granular zone were seen in apposition to the nuclear envelope(Fig. 6).

Chromosomes at metaphase 1 and nucleolar reformation at the dyad stage

Meiotic metaphase chromosomes showed a homogenous fibrous appearance (Fig. 7)with individual fibrils between 6 and 15 nm in diameter (Fig. 8). During chromosomedecondensation at the time of the newly formed dyad, numerous fibrous bodies wereseen adjacent to the chromosomes and free in the nucleoplasm (Fig. 9). As deconden-sation progressed these bodies decreased in number and increased in size (Figs.10-12), and were seen to be composed of fibrils 5-10 nm in diameter (Fig. 12). Thesepre-nucleolar bodies coalesced on to the nucleolus organizer forming the main nucleoluswhich was composed of a fibrillar mass with a lightly staining zone on the surface(Figs. 13, 14), essentially similar to the inactive nucleoli described in Helianthustuberosus (Jordan & Chapman, 1971).

Figs. 4-6. Various aspects of nucleolar dissolution.

Figs. 4, 5. Pachytene-diplotene, showing the fragmentation of the granular andfibrillar region of the nucleolus. Fig. 4, x 16500; Fig. 5, x 9500.

Fig. 6. Diplotene. A large granular region fragment is seen in apposition to theremnant nuclear envelope, x 12750.

n 6 B. T. Luck and E. G. Jordan

m

The nucleohis during microsporagenesis 117

Chromosomes at anaphase 2 and nucleolar reformation at the tetrad stage

Anaphase 2 chromosomes showed loosely arranged fibrils 6-15 nm in diameter(Fig. 15). Nucleolar reformation at the tetrad stage occurred in a similar fashion to thatat the dyad stage, with the appearance of pre-nucleolar bodies which coalesced on tothe nucleolus organizer (Fig. 16) to form the main nucleolus of the young tetrad,which was also composed of a large fibrillar mass with a lightly staining zone on thesurface (Fig. 17). When primary exine development was complete the young micro-spores were seen to have nuclei with a few areas of condensed chromatin surroundedby large areas of dispersed chromatin. The lightly staining zone of the nucleolus was thenseen in the interior of the fibrillar region. Concurrent with this rearrangement was theappearance of a peripheral granular region, and the production of some nuclear bodies(Fig. 18). Such nucleoli had the same structural characteristics as those of artichokenuclei with activated ribosomal cistrons (Jordan & Chapman, 1971).

DISCUSSION

During meiotic prophase in Endymion non-scriptus the nucleoli undergo segregationof their different zones with the appearance of the organizer as a discrete mass on thesurface of the fibrillar region (Jordan & Luck, 1976). Two separate nucleoli will bebrought together by the synapsing of their organizers. This explains how only asingle large nucleolus is seen at pachytene.

The accessory nucleoli seen at this stage have an ultrastructure similar to thefibrillar zone, sometimes denser, and can be readily distinguished from the chromatinby the diameter of their constituent fibrils. They are not thought to be a condensedpart of the nucleolus organizer (La Cour & Wells, 1975). Although the nucleolusorganizer may eventually condense and appear as darkly stained as the rest of thechromatin, the time when this happens may vary with species. In Endymion non-scriptus the organizer is pale-staining at pachytene, a condition also seen in Liliumlongiflorum (Williams, Heslop-Harrison & Dickinson, 1973), but in Allium cepa thenucleolus organizer is only partially lightly stained (Esponda & Gimenez-Martin,

Figs. 7—10. Metaphase 1 chromosomes and the appearance of pre-nucleolar bodiesduring chromatin decondensation.

Fig. 7. Metaphase 1 chromosomes show a homogeneous fibrillar appearance. Thebroken down nuclear envelope forms a discontinuous membrane around the spindlearea. A few cytoplasmic organelles are seen in the spindle area, x 5100.

Fig. 8. Higher magnification of a metaphase 1 chromosome which shows it to becomposed of fibrils between 6 and 15 nm in diameter, x 31000.

Fig. 9. Newly formed dyad. Prenucleolar bodies are adjacent to the decondensingchromosomes and free in the nucleoplasm. A nucleoloid is present in the cytoplasm.Nuclear pores (arrowheads), x 10000.

Fig. 10. A later stage of the dyad as judged by the further decondensation of thechromosomes and the larger pre-nucleolar bodies. A polysynaptonemal complex ispresent in the nucleoplasm. x 10000.

n 8 B. T. Luck and E. G. Jordan

The nucleolus during microsporogenesis 119

1975). We have not seen lightly staining zones in stages later than pachytene-diploteneuntil the times of nucleolar reformation after both divisions, when the nucleolusorganizer has the characteristic lightly stained appearance.

During diplotene desynapsis the separated fibrillar and granular zones of thenucleolus become fragmented. The chromosomes become very diffuse at pachytene-diplotene, and the remaining fragments of the nucleolus disappear. It is difficult todecide upon the fate of the dispersed regions of the nucleolus. However, it is con-ceivable that some ribosomal precursor fibrils will become part of the matrix ofcondensing chromosomes (Lafontaine & Chouinard, 1963; Lafontaine, 1958). Thismatrix can be considered to be arrested nucleolar machinery, in particular its highmolecular weight ribosomal RNA precursors associated with the metaphase chromo-somes (Fan & Penman, 1971). The consequence of this would be that some nucleolarformation could occur without any requirement for concurrent synthesis. Althoughsome RNA may be synthesized during telophase in mitotic divisions (Prescott &Bender, 1962; Monesi, 1964), evidence shows that pre-nucleolar bodies can be formedin the absence of RNA synthesis (Stevens & Prescott, 1971; Stockert et al. 1970;Phillips, 1972; Gimenez-Marti'n et al. 1974; Semeshin et al. 1975; Risueno et al.1976) or protein synthesis (Stockert etal. 1970; Gimenez-Marti'n et al. 1974). Further,even under conditions of normal protein synthesis the reappearance of nucleoli at theend of mitosis involved the utilization of proteins which were synthesized beforemitosis (Harris, 1961). There is thus good evidence that some of the remnants ofthe dispersed prophase nucleolus are reutilized in nucleologenesis.

In anaphase and telophase of mitosis a fibrillo-granular material has been reportedcoating the chromosomes. Whether or not such a fibrillo-granular material is a stage innucleolar reformation has been a matter of some debate. The material has even beencalled the pre-nucleolar material. But because such a material has clearly definedgranules which do not appear in the first recognizable nucleolar precursor bodies ithas been argued that it is not nucleolar precursor material (Chouinard, 1966, 1971).

In this work on Endymion non-scriptus at meiosis no material having the precisenature of a fibrillo-granular coating of the chromosomes was seen. However, the areasbetween the decondensing chromosomes do show the presence of scattered granuleswhich may be its counterpart, and it is possible that a stage showing a fibrillo-granularcoating of the chromosomes occurs but is of short duration and has therefore not been

Figs. 11—14. Nucleologenesis at the dyad stage.

Fig. 11. Later stage of the dyad than Figs. 9, io, as judged by chromatin dispersal,fewer and larger pre-nucleolar bodies, arrowheads, nuclear pores, x 10000.

Fig. 12. Higher magnification of outlined rectangular area of Fig. 11, showing thepre-nucleolar bodies to be composed of fibrils 5-10 nm in diameter, x 39000.

Fig. 13. Large fibrillar mass adjacent to 2 lightly staining nucleolus organizerregions which are continuous with more darkly staining condensed chromatin. Apre-nucleolar body is present in the nucleoplasm. x 17250.

Fig. 14. Main nucleolus of the dyad composed of a large fibrillar region with alightly staining organizer region embedded in its surface, x 14500.

120 fi. 7\ Luck and E. G. Jordan

The nucleolus during microsporogenesis 121

observed. Our evidence would argue against any involvement of a granular material innucleologenesis, the fibrillar pre-nucleolar bodies appearing in scattered placesamongst the chromosomes in a manner consistent with an origin from some fibrillarchromosomal matrix.

The appearance of a fibrillo-granular mass in mitosis might well be the expressionof non-nucleolar RNA synthesis in such a scheme of gene reprogramming as proposedby Goldstein (1976) and may then be quite different for meiosis. Whether the pre-nucleolar bodies and nucleoli arise from a chromosomal fibrillar matrix directly(Lafontaine & Lord, 1974) or whether the so called pre-nucleolar material aroundanaphase chromosomes must be an intermediate cannot yet be settled, but a moredetailed investigation of the 2 meiotic anaphases could help to clarify the situation.

The collection of the pre-nucleolar bodies into one nucleolus, the process ofnucleolar reorganization, has been shown to be dependent on the presence of anucleolus organizer. In the absence of organizers the many pre-nucleolar bodies,though undergoing some enlargement and perhaps fusion, remain separate or 'un-organized' (Swift & Stevens, 1966). From the use of inhibitors it has been concludedthat nucleolar organization is dependent on RNA synthesis (Risueno et al. 1976).

Our observations show the effect of the organizer in marshalling the pre-nucleolarbodies to form the large fibrillar mass of the newly reformed nucleolus. The externalcondensed organizer of the newly reformed nucleolus has a strong resemblance to thestructure of an inactive organizer (Jordan & Chapman, 1971, 1973), suggesting thatnucleolar reorganization can be performed by an organizer that is not synthesizingrRNA. Nucleolar reformation in the absence of rRNA synthesis is in line with theactinomycin D experiments of Semeshin et al. 1975. Perhaps there is some force ofattraction between the nucleolus organizer and nucleolar material which is notdependent on RNA synthesis. A further indication that a force of 'attraction' existsbetween the organizer and the nucleolar material even when the organizer is notactive in synthesis is the fact that the fibrillar regions of nucleoli remain closelyadherent to withdrawn condensed inactive organizers. However, the earlier dispersedlocation of the fibrillar pre-nucleolar bodies is difficult to explain if it is argued thateven a synthetically inactive organizer could suffice for nucleolar reorganization,because such exists at the time when these structures first appear. Although nucleolarorganization may not require synthesis or extensive decondensation of the organizer

Fig. 15. Anaphase 2 chromosome composed of loosely arranged fibrils 6-15 nm indiameter, x 17250.Fig. 16. Pre-nucleolar bodies of the young tetrad. A pre-nucleolar body has co-alesced on to the lightly stained, condensed chromatin region (nucleolus organizer).x 17250.Fig. 17. Newly reformed nucleolus of the young tetrad composed of a large fibrillarmass with the nucleolus organizer on the surface. Nucleolar vacuoles are presentwithin the fibrillar region, x 21 375.Fig. 18. Nucleolus of young microspore with the nucleolus organizer in the interiorof the fibrillar region which is surrounded by granular regions. A spherical nuclearbody is on the surface of the fibrillar region, x 19125.

122 B. T. Luck and E. G. Jordan

chromatin, it might be argued that at least some chromatin from the apparentlycondensed organizer had begun to decondense in the reorganization process, eventhough most of it could still be recognized as a condensed structure, the lightlystaining zone. Further work on the amount of RNA synthesis associated with con-densed organizers especially at the end of anaphase I of meiosis could help to settlethe question.

The dispersion of the nucleolus organizer into the nucleolus is a process whichaccompanies the arrival of a granular zone (Jordan & Chapman, 1971, 1973)- Theoutcome of nucleologenesis at the 2 meiotic divisions is different in this respect.While at the dyad a nucleolus is organized at the condensed organizer, it remains apurely fibrillar structure not developing a granular zone, while at the tetrad stage thenucleolus organizer moves into or becomes engulfed by the enlarging nucleolus as thegranular zone appears. If the granules are evidence of resumed nucleolar functionthen nucleologenesis in the first division occurs without any resumption of RNAsynthesis or processing, while at the end of the second meiotic division nucleolo-genesis results in both the resumption of rRNA synthesis and processing. This isconsistent with Taylor's (1958) interpretation of his experiments on the time of RNAsynthesis in Tulbaghia. It is therefore possible to speculate that the nucleolus organiz-ing region may collect a fibrillar zone while remaining largely in a condensed con-figuration, but that its dispersal must precede or accompany any restoration of RNAsynthesis or processing.

REFERENCES

CHOUINARD, L. A. (1966). Nucleolar architecture in root meristematic cells of AIHum cepa.Natrt. Cancer Inst. Monogr. 23, 125-143.

CHOUINARD, L. A. (1971). Behaviour of the structural components of the nucleolus duringmitosis in A Ilium cepa. In Advances in Cytopharmacology, vol. 1 (ed. F. Clementi and B.Ceccarelli), pp. 69-87. New York: Raven Press.

CHOUINARD, L. A. (1975). An electron-microscope study of the intranucleolar chromatin duringnucleologenesis in root meristematic cells of Allium cepa. J. Cell Sci. 19, 85-102.

ESPONDA, P. & GIMENEZ-MARTI'N, G. (1975). Nucleolar organiser ultrastructure in Allium cepa.Chromosoma 52, 73-87.

FAN, H. & PENMAN, S. (1971). Regulation of synthesis and processing of nucleolar componentsin metaphase-arrested tells. J. molec. Biol. 59, 27—42.

GIMENEZ-MARTI'N, G., DE LA TORRE, C, FERNANDEZ-GOMEZ, M. E. & GONZALEZ-FERNANDEZ, A.(1974). Experimental analysis of nucleolar reorganization. J. Cell Biol. 60, 502-507.

GOLDSTEIN, L. (1976). Role for small nuclear RNAs in 'programming' chromosomal infor-mation? Nature, Lond. 261, 519-521.

HARRIS, H. (1961). Formation of the nucleolus in animal cells. Nature, Lond. 190, 1077-1078.JORDAN, E. G. & CHAPMAN, J. M. (1971). Ultrastructural changes in the nucleoli of Jerusalem

artichoke {Helianthus tuberosus) tuber discs. J. exp. Bot. 22, 627-634.JORDAN, E. G. & CHAPMAN, J. M. (1973). Nucleolar and nuclear envelope ultrastructure in

relation to cell activity in discs of carrot root (Daucus carota L.). J. exp. Bot. 24, 197-209.JORDAN, E. G. & GODWARD, M. B. E. (1969). Some observations on the nucleolus in Spirogyra.

J. Cell Sci. 4, 3-15.JORDAN, E. G. & LUCK, B. T. (1976). The nucleolus organizer and the synaptonemal complex in

Endymion non-scriptus (L.). J. Cell Sci. 22, 75-86.LA COUR, L. F. & WELLS, B. (1975). The nucleolus at prophase of meiosis in three plants: an

ultrastructural study. Proc. R. Soc. B 191, 231-243.

The nucleolus during microsporogenesis 123

LAFONTAINE, J. G. (1958). Structure and mode of formation of the nucleolus in meristematiccells of Viciafaba and Allium cepa.J. biophys. biochem. Cytol. 4, 777-784.

LAFONTAINE, J. G. & CHOUINARD, L. A. (1963). A correlated light and electron microscopestudy of the nucleolar material during mitosis in Viciafaba. J. Cell Biol. 17, 167-201.

LAFONTAINE, J. G. & LORD, A. (1974). A correlated light- and electron-microscope investigationof the structural evolution of the nucleolus during the cell cycle in plant meristematic cells(Allium porrum). J. Cell Sci. 16, 63-93.

MCCLINTOCK, B. (1934). The relationship of a particular chromosome element to the develop-ment of the nucleoli in Zea mays. Z. Zellforsch. mikrosk. Anat. 21, 294-328.

MONESI, V. (1964). Ribonucleic acid synthesis during mitosis and meiosis in the mouse testis.J. Cell Biol. 22, 521-532.

MORENO DIAZ DE LA ESPINA, S., RISUENO, M., FERNANDEZ-GOMEZ, M. & TANDLER, C. J.(1976). Ultrastructural study of the nucleolar cycle in meristematic cells of Allium cepa.y. Microscopie Biol. cell. 25, 265-278.

NOEL, J. S., DEWEY, W. C, ABEL, JR., J. H. & THOMPSON, R. P. (1971). Ultrastructure of thenucleolus during the Chinese hamster cell cycle. J. Cell Biol. 49, 830-847.

PHILLIPS, S. G. (1972). Repopulation of the postmitotic nucleolus by preformed RNA. J. CellBiol. 53, 611-623.

PICKETT-HEAPS, J. D. (1970). The behaviour of the nucleolus during mitosis in plants. Cytobios6, 69-78.

PRESCOTT, D. M. & BENDER, M. A. (1962). Synthesis of RNA and protein during mitosis inmammalian tissue culture cells. Expl Cell Res. 26, 260-268.

RISUENO, M., MORENO DIAZ DE LA ESPINA, S., FERNANDEZ-GOMEZ, M. & GIMENEZ-MARTI'N,G. (1976). Ultrastructural study of nucleolar material during plant mitosis in the presence ofinhibitors of RNA synthesis. J. Microscopie Biol. cell. 26, 5—18.

SEMESHIN, V. F., SHERUDILO, A. I. & BELYAEVA, E. S. (1975). Nucleoli formation under inhibi-ted RNA synthesis. Expl Cell Res. 93, 458-467.

STEMPACK, J. G. & WARD, R. T. (1964). An improved staining method for electron microscopy.y. Ceil Biol. 22, 697-701.

STEVENS, A. R. & PRESCOTT, D. M. (1971). Reformation of nucleolus-like bodies in the absenceof postmitotic RNA synthesis, y. Cell Biol. 48, 443-454.

STOCKERT, J. C, FERNANDEZ-GOMEZ, M. & GIMENEZ-MARTI'N, G. (1970). Organization ofargyrophilic nucleolar material throughout the division cycle of meristematic cells. Proto-plasma 69, 265-278.

SWIFT, H. (1959). Studies on nucleolar function. In A Symposium on Molecular Biology (ed.R. E. Zirkle), pp. 266-303. Chicago: University of Chicago Press.

SWIFT, H. & STEVENS, B. J. (1966). Nucleolar-chromosomal interaction in microspores of maize.Natn. Cancer Inst. Monogr. 23, 145-167.

TAYLOR, J. H. (1958). Incorporation of phosphorus-32 into nucleic acids and proteins duringmicrogametogenesis of Tulbaghia. Am. y. Bot. 45, 123-131.

WILLIAMS, E. J., HESLOP-HARRISON, J. & DICKINSON, H. G. (1973). The activity of the nucle-olus organising region and the origin of cytoplasmic nucleoloids in meiocytes of Lilium.Protoplasma 77, 79-93.

{Received 26 August 1976)