THE EFFECTS OF LEAF PRIMORDIA ON DIFFERENTIATION IN THE STEM

Transcription

1 [ 445 ] THE EFFECTS OF LEAF PRIMORDIA ON DIFFERENTIATION IN THE STEM BY B. S. YOUNG {Received 23 October 1953) (With 24 figures in the text) (i) INTRODUCTION The problem to be studied here is, what are the physiological causes leading to the differentiation of cambiform cells in the plant axis and their aggregation into desmogen strands, and how are the leaf primordia concerned in the process? There is indeed a further problem concerning the differentiation of the desmogen strands into xylem and phloem but it seems to the author that this latter problem must wait until later. Since so little work had been done on the plant embryo, it was decided to experiment on the shoot apex. A reading of the literature concerning the differentiation of conducting strands in higher plants led to the following conclusions: (1) It seems likely that there is in the apices of many plant species a ring of tissue physiologically differentiated, within which are differentiated subsequently the precursors of the vascular system the desmogens (see Helm, 1931, 1932). (2) The apical meristem is the main agent in the further organization of the shoot. It determines most, if not all, of the distributions of tissues in the axis. (3) The leaves determine their traces, at least in Lupinus albus. Yet the work of Wardlaw (1945 to 1950) indicates that in ferns and in Primula polyantha the stele is largely cauline. (4) Helm (1932) found that in Ricinus and Lysimachia removal of a young leaf primordium, which was sometimes the youngest visible, caused its trace to dedifferentiate partially. Hence differentiation of a leaf-trace appeared to depend, partly at least, on its leaf. (5) The nature of this induction of a trace by a young leaf is quite unknown. (2) MATERIAL AND METHODS The experiments reported were performed on Lupinus albus since the plant was easy to grow in large numbers at all times of the year, and the apex was found to be of fair size and was easily uncovered for the experiments. Also when the leaves were removed the leaf stumps did not bleed so much as to obscure the apex, and the plant had been used very often by other workers (e.g. Ball (1949); Snow & Snow, 1931, 1947), so that their results could be compared with any from this investigation and used if needed. It was found that the lupin apex was very robust, and few of the apices died after the operations. The seeds were soaked in tap water for hr. and were then sown separately in pots: the pots were kept in a warm greenhouse in which it was found that the plants

2 446 B. S. YOUNG grew satisfactorily all the year round. They were usually ready for use about a fortnight after sowing. If left any longer, they tended towards the flowering state. The apices were exposed quite easily by bending back the older leaves, and usually by removing some of the younger ones also. A binocular dissecting microscope was used for the last stages of exposing the apices and for the operations. Care was taken throughout that the apex should not be desiccated, the light being filtered through heat-absorbent glass. Primordia were removed by means of small knives which were kindly made by the chief technician of the Clarendon Physical Laboratory. After the operations the plants were replaced in the greenhouse for periods of from 3 to 21 days. All experiments were run in pairs, one plant being used for the experiment, and the other for control. So far as possible the members of a pair were grown under similar conditions. The apices were fixed in 70% ethyl alcohol and embedded in paraffin wax of melting-point about 55 C. It was found that safranin and fast green were the most useful and effective of the stains tried, although in some later experiments Chlorazol Black E was used. This last was very efficient in differentiating cell walls, but not as good as safranin and fast green in differentiating the cell contents. The embedded apices were sectioned with a microtome, the sections being cut 10 fj. thick (occasionally 20 fx), stained, and mounted in Canada balsam. After removing one primordium it was found useful to arrange to identify it later by splitting with a cut another numbered primordium. Thus after removal of P^, P3 was usually so split. There were few failures during the operations or during the later treatment. Failures in the operations were commonly due to too large or severe a cut in the removal of a primordium, or to removing many primordia successively. A few experiments were rejected owing to a change in the apex to the flowering condition. This change was easily seen. The terminology used is the following. Cambiform cells are prosenchymatous elements, densely cytoplasmic and having a length to breadth ratio {LIB) of from 2-5 to 3-5. This ratio is defined as the ratio of the greatest to the least visible dimensions of the cell. Usually the ratio was determined for the dimensions seen in transverse sections, although it can be determined in longitudinal sections also. By this definition no developmental series is implied as it refers only to anatomical characteristics: thus it is not liable to abuse, as is possible with the terms procambium and prodesmogen. Merismatic cells are defined as isodiametric cells, densely cytoplasmic, with conspicuous nuclei, and with an LjB ratio of from i-o to 1-5. The leaf primordia are numbered according to the terminology of Snow & Snow (1931). The primordia visible at the time of operation are called P^,P^,P^, etc., in ascending order of age. Primordia arising after an operation are called 1^,1^,1^, etc., in descending order of age. (3) THE NATURE OF THE OPERATIONS The operations performed on L. albus apices consisted entirely of the removal of very young leaf primordia. From Helm's observations (1932) and those of Snow & Snow (1947) on the determination of leaves and leaf traces, it seemed probable that the removal of a leaf-primordium would lead to the reduction of its leaf traces to a greater or lesser degree.

3 The effects of leaf primordia on differentiation in the stem 447 (4) THE NORMAL APEX OE LUPINUS ALBUS The first step was to investigate the normal apex of L. albus. Ball (1949) had carried out a thorough investigation of the normal lupin apex, and the results described below to a large extent confirmed his results. Twelve plants were grown and when they had reached the requisite stage the apical half-inch of each was cut off and fixed. After embedding, six apices were sectioned longitudinally and six transversely. Both sets were stained with safranin and fast green. The phyllotaxis of the lupin apex has been described by Snow & Snow (1931, 1947)- It is a spiral with leaf contacts i, 2 and 3, and mean divergence angle of The spiral may run either clockwise or anticlockwise; the latter was found to be the commoner. Serial transverse sections showed that the apical 40-50^ consist wholly of merismatic cells, staining deeply with safranin. Within this mass no evidence could be found of anatomical cellular differentiation. In order to demonstrate this objectively, cell outlines were drawn on graph paper, on which they were projected through a microscope. Measurements were made of the greatest dimension L, and of the least dimension, B. The ratio L/fi was then computed, and this value plotted against cell number (Table i). Table i. The numbers of cells with various length-breadth ratios. Cells from apical meristem LIB ratio No. of cells The resulting curve is unimodal, pointing to the presence of only one cell type in the tissue. LIB for this tissue was i-o-i-5. This result, which was obtained both from longitudinal and transverse sections, may be criticized on the grounds that in transverse section a similar result would be given by parenchyma; but the merismatic cells were distinguished from parenchyma by not having vacuoles and by their dense protoplasmic contents. Leaf I arises about 20-30/x below the extreme apex, and to accommodate the primordium the stem section changes from circular to oval, the narrow end of the oval being towards the primordium (see Fig. 3). This change is due to the formation of the foliar buttress of the primordium (Louis, 1935). At this level there is no sign of vacuolation in the cells of the medulla. Ball (1949) thinks that this is against the existence of a meristem ring, distinct from the medulla, in Lupinus. But there is no reason why the ring, if present, should always reach up to the level of the first leaf. Neither transverse sections (Fig. 3) longitudinal (Fig. 10) revealed any cambiform cells at this level. Ball states that 'the procambium occurs below each foliar primordium at about the same time that the new periclinal divisions in the second tunica layer initiate the primordium' (1949, p. 441). He further claims that the procambium below a very young primordium is unconnected with that in the stem, and that this suspended trace develops basipetally and basifugally. In the apices examined by the author no evidence of this could be found. Leaf 2 arises on the stem approximately 50 fx below the apex. At this level the cells in the middle of the stem first appear vacuolated (Fig. 4). It is difficult to give an

4 448 B. S. YOUNG Fig- 3 Fig-4 Fig-5 Fig. 7 Fig-8 Figs Transections of the normal shoot of Lupinus albus, down to the insertion of leaf 4, showing the merismatic cells (shaded), and the parenchymatous medulla and cortex (not shaded). Desmogen strands are shown black. All x 32.

5 The effects of leaf primordia ort differentiation in the stem 449 objective quantitative measure of these differences, but the medulla cells stain less deeply than the cells of the cortex. The second primordium is still wholly merismatic, although its abaxial cells appear to be a little larger than the rest. There was no sign of cambiform tissue just below the second leaf, except in three apices which were in a late stage of the plastochron. Leaf 3 arises 70-80/t below the apex (Fig. 6). At this level the medulla is well marked, and a cortex of parenchyma extends about one-third of the way around the stem. Most of the band of cells between the medulla and the cortex differs little from the merismatic cells found in the apex, in that L/5 approximates to i. But with an oil-immersion objective it is possible to distinguish cells in this band or ring which are more elongated in transverse section than are the majority. For these cells L/5 is 2-5. On a graph showing LIB against cell number (Table 2) two modes appear on the curve, indicating the existence of two cell types in the tissue at this level, the more extended cells being cambiform. These cells show up well in longitudinal section (Fig. 10), and they occur in the arc of the stem occupied by the third leaf. The shorter cells are still merismatic. Table 2. Frequencies of different LIB ratios of the cells in the meristem ring at the level of P^ LIB ratio I-I 'O 'o No. of cells The apex of the third leaf consists wholly of merismatic cells (Fig. 10). Further down the leaf there appears an abaxial band of parenchyma cells, and also lower down an adaxial band. The abaxial band continues down to complete the cortex of the stem, as was noted by Louis (1935) ^ below the apex there appear three groups of cambiform cells, or desmogens, below the third leaf. In its foliar buttress there is a well-marked central compact cylinder of cambiform cells (Fig. 7), which is a desmogen strand, the precursor of the median trace of that leaf. Its two lateral strands are more weakly developed. A little below this level the fourth leaf joins the axis (Fig. 8). It has three well-marked strands, but still no sign of differentiated xylem elements or phloem. Below the level of the fourth leaf there is present in the stem a ring of tissue consisting of desmogen strands separated by bands of mixed cambiform and merismatic cells (Fig. 8). The desmogen strands can be traced up the stem, and are found to connect with the third and fourth leaves, or with older leaves. Ball (1949) states that the bands between the traces differentiate into phloem and vascular cambium later (see Ball, 1949, figs. II, 13, and p. 441). Longitudinal sections support these observations by showing that the first cambiform cells are found in the buttress of the third leaf. Above them there is only a mass of merismatic cells, with some parenchymatous medulla, but no cortex (Figs. 9, 10). The cambiform cells are distinctly elongated and narrow. In the medulla are fairly wellmarked file meristems (Fig. 9). The cambiform cells of the apex are everywhere continuous with more cambiform cells below and with merismatic cells above. There was no evidence in the apices examined

6 450 B. S. YOUNG of the suspended traces claimed to be present by Ball (1949); and it seems to the writer that the two figures given by Ball (1949, figs. 5, 6) are not very convincing. Examination of the lower parts of the stem shows that the median and cathodic traces Fig. 10 Fig. 9. Longisection of normal Lupinus apex in the median plane of leaf i, showing the merismatic regions (shaded) and the desmogen strands and cambiform cell areas (cross-hatched). Also shown are the stem apex {A), the cortex (C), the medulla (M) and the file meristems of the medulla (F). Leaves i to 3 are numbered, x 80. Fig. 10. Longisection through margin of the base of the third leaf of normal Lupinus apex, showing merismatic cells (M), cambiform cells of desmogen strand (C), medullary parenchyma {MP), cortical parenchyma (CF) and merismatic cells of the upper marginal part of leaf 3. The broken line indicates the outline of leaf 3 in median section, x 320. of each leaf fuse with traces from the leaf three plastochrons older, whilst the anodic trace fuses with the cathodic trace of the leaf two plastochrons older (Snow & Snow, 1947,fig. i)- In summary it may be said that the apex of L. albus consists wholly of merismatic cells. The leaf primordia arise in a regular spiral sequence, and the youngest two are also

7 The effects of leaf primordia on differentiation in the stem 451 wholly merismatic. At the level of insertion of the second leaf there appears a ring of tissues extending to the periphery, which consists entirely, or nearly entirely, of merismatic cells and can be distinguished from a central medulla of vacuolated cells. Within the ring elongated cambiform cells first appear at the level of insertion of the third leaf, or of the second at late plastochron. At the level of insertion of the third leaf the cortex also begins to be formed part by part, but it is completed only at the level of the fourth or fifth leaf. It is continuous with the blocks of abaxial parenchyma of the leaves above and surrounds the merismatic ring. Still a third change appears at the level of the third leaf, for here it can first be seen that the cambiform cells in the merismatic ring are aggregated into a desmogen strand, the median trace of the third leaf. The two lateral traces of this leaf are fainter. The older leaves have clearly marked median and lateral traces at their insertions, and still lower down the remainder of the merismatic ring between the traces is converted into cambiform cells, which finally give rise to xylem and phloem. Lastly the effects of hydrogen peroxide were tried, 2% hydrogen peroxide being dropped on to fresh sections of the stem about 8o-ioo/it below the apex. It caused considerable effervescence, and if most of the liquid was carefully removed with filterpaper there was found to be a dark ring with a silvery mass on each side of it. Microscopic examination showed that this dark ring consisted of cells which were not causing the evolution of oxygen from the solution, whereas the cells on each side were doing so. This agrees with observations by Helm (1931). It was also found that sections from much lower down the stem gave similar results. However, when the fresh section was cut above the level of the first leaf, no evolution of oxygen occurred at all. From this it was concluded that the merismatic cells of the apex do not cause the evolution of oxygen from hydrogen peroxide, nor do those which form the stele at any stage in their development. Van Fleet (1950) has shown that the latter cells have only weak oxidizing power, but what this may imply is unknown. (5) EXPERIMENTAL RESULTS The main results of the removals of P^, are summarized in Table 3. They were all observed in transverse sections of the shoot apices, which were compared with sections of unoperated plants as controls, made at the corresponding levels of the axis. These levels were identified by means of the insertions of the successive leaves, as will be made clear below. Only the internodes below the levels of insertion of the removed primordia will be described as it was found that it was only these that were changed anatomically by the operations. Table 3. Tissues in the arc of the stele of the removed leaf-primordium Pg Length of expt. Normal control apices Apices with Pj removed Apices with P^ removed, and stump auxinated 3 6 days M+C M + ip) M 6 9 days (M) + C + {D) {M) + P M 9 12 days C + D P M 12 days and more D (xylem and phloem) P M M=merismatic cells; C = cambiform cells; P=parenchyma; Z) = desmogen strands. Letters in brackets mean that the tissue is not always present, or is scanty.

8 452 B. S. YOUNG Expt. I. Po removed, experiment lasting 3 ^o 6 days Twelve plants were treated in this manner. In the short period of the experiment usually one new leaf (/j) had arisen: in two plants /g was found to have arisen also. So in judging the results from the sections comparisons were made between the internode below P.^ in the experimental plants, and that below the third or fourth leaf in the controls. It was found that in ten of the experimental plants there was no formation of cambiform cells in the sector of internode directly below the removed primordium. In the other two there had been a very slight formation of cambiform cells in this sector. But in four of these plants which were kept for 6 days, and in one which was kept for 4 days, there was observed a slight formation of parenchyma in the arc of the stele below the stump of P.^ (Fig. 11). By comparison, in the controls there was a much more extensive development of cambiform cells below the third leaf, and no parenchyma had been formed in the stelar ring. These results clearly indicate that in the absence of the primordium there is no significant development of cambiform cells in the sector of the stelar ring just below it, but parenchyma begins to be formed after 4 days. Expt. 2. P^ removed, experiment lasting 6 to g days Twelve plants were treated in this manner. In these plants two or three new leaves arose during the experiment, or in one of them four new leaves. So comparisons were made between the internode below P^ in the experimental plants and that below the fourth or fifth leaf in the controls. In all the twelve experimental plants the results were similar. Below the stump of the removed P., there was a gap in the stelar meristem ring of the axis. This gap consisted only of small parenchyma cells in nine plants, and in the other three plants of small parenchyma cells interspersed with a few groups of merismatic cells remaining amongst them (Fig. 12). The controls showed at the corresponding levels considerable numbers of cambiform cells, which were aggregated into desmogen strands. In these plants no parenchyma was developed except as a sheath around each bundle. This result is more definite than that of Helm (1932), who found a weak development of traces below the stumps of removed primordia in Ricinus. Expt. 3. P.2 removed, experiment lasting g to 12 days During the period of this experiment, in which twelve plants were used, from three to five new leaves arose on the apex. Comparisons between the experimental plants and the corresponding levels in the controls showed in the former a gap completely filled with parenchyma in the sector of the stelar ring below the removed primordium (Fig. 13). This gap was bordered by normal traces leading to higher leaves and by interfascicular phloem and cambium similar to those found in the controls. The parenchyma cells in the gap were strongly vacuolated and as large as the cells of the cortex (Fig. 14). They gradually merged with the rather larger cells of the medulla. Expt. 4. P., removed, experiment lasting more than 12 days The results from the twelve plants of this experiment were closely similar to those from the 9- to 12-day experiment. There was a well-marked parenchymatous gap in the arc of the stele below P^ (Fig. 15), and although three of these plants were grown for

9 The effects of leaf primordia on differentiation in the stem 453 Fig- 12 Fig. II. Transaction of a Lupinus apex 4 days after removal of P^, showing the start of the formation of parenchyma (P) in the sector of the stele below the removed primordium. Wound tissue and necrotic divisions are marked with black shading and short lines respectively. M, medulla, x 48. Fig. 12. Transection of a Lupinus apex 8 days after removal of P2, made at level of insertion of P3, showing a gap in the stele below P^ with groups of merismatic cells (M) remaining in the gap. x 32. Fig. 13. Section of an apex 10 days after removal of Pj, showing a complete parenchymatous gap in the stelar ring below Pj. x 40. Fig. 14. Enlargement of the parenchymatous cells in the stele shown in Fig. 13. x 320. Fig. 15. Section of a Lupinus axis 21 days after removal of Pj, showing a well-marked parenchymatous gap in the sector of stele below Pj. New Phytol. 53, 3 29

10 454 B. S. YOUNG 21 days, the cells of the gap did not develop further. In all these experiments the only part of the stele affected by the removal of the young primordium was the sector of the stele below the stump of that primordium. The tissues in other parts of the stem and stele were similar to those of the normal control plants. Wardlaw (1950), dealing with results obtained from experiments on the removal of primordia from Primula, states: 'In some instances, where rather deep incisions had been made in removing the leaf primordia, parenchymatous gaps were present in the vascular tissue. These gaps were due to the formation of wound parenchyma from procambial tissues.' In the present experiments the depth of the cuts by which the primordia had been removed was carefully noted by comparing the projected images of the transverse sections of the experimental plants with a set of drawings of normal apices. In this way it was possible to see that usually none of the cortex had gone, and often not all of the primordium had gone, so that the outermost tips of its stipules remained. Yet the results all agreed. For even in four apices where rather too large a cut seemed to have been made, it was found that the cortex outside the stelar ring was of uniform width all round the stem. Hence in the present experiments the gaps below the removed primordia cannot have been due to the formation of wound parenchyma. These results clearly indicate that the presence of the young primordium is necessary for the differentiation and development of the leaf traces below it. Now it is known that the young leaf is an active producer of auxin (Snow, 1929; Goodwin, 1937), and hence the writer thought that this auxin might be the main factor causing the traces to differentiate. So to test this hypothesis, experiments were next made in which a supply of indoleacetic acid, as an auxin, was added to the stump of the removed primordium in lanoline. The lanoline was prepared by mixing equal parts of water and anhydrous wool fat, and contained one part in a thousand of indole-acetic acid. The experiments were otherwise similar to those described above. Expt. 5. P2 removed and its stump auxinated: experiment lasting 2 to 6 days Twelve plants were treated in this and the following experiments as before, and similar sets of untreated plants were run as controls. In ten plants of this experiment two new leaves arose after the operation, and in the other two plants only one leaf. On comparing the experimental plants and the controls in transverse sections, it was found that no cambiform cells had developed below the auxinated stump of Pj, but instead there seemed to be more abundant merismatic cells than in the controls (Fig. 16). This was ascribed to the retention of the merismatic condition under the infiuence of auxin. In other parts of the stele in the experimental plants, for instance in the sector below Pj, cambiform cells had developed as in the sector below the third leaf in the controls. Expt. 6. P, removed and its stump auxinated: experiment lasting 6-9 days In eleven of these twelve experimental plants, four leaves arose in the course of the experiment, and in transverse sections it was seen that in the sector of the steele below the auxinated stump there was a band of merismatic cells still unbroken by parenchyma and also without cambiform cells (Fig. 17). In the controls at the corresponding level well-marked desmogen strands had developed.

11 The effects of leaf primordia on differentiation in the stem 455 Fig. 16. Section of a Lupinus apex with P^ removed and stump auxinated, after 4 days, showing a slight thickening of the sector of the merismatic ring below P^. This sector contained no cambiform cells. Fig. 17. Section of a Lupinus apex with Pj removed and stump auxinated, after 8 days, showing in the sector of the ring below P^ a merismatic band containing no desmogens nor cambiform cells, x 40. Fig. 18. Section of a Lupinus apex with Pz removed and stump auxinated, after 10 days, showing a band formed of merismatic cells alone in the sector below P^. x 40. Fig. 19. Enlargement of part of the band below P^ shown in Fig. 18. MB, merismatic band. 29-2

12 456 B. S. YOUNG Expt. 7. P., removed and its stump auxitiated: experiment lasting g to 12 days or more The twelve plants so treated did not differ much from those of the 6- to 9-day experiment. In the sector of the stele below the auxinated stump there was a well-marked band of merismatic cells, showing no sign of differentiation and unbroken by parenchyma (Figs. 18, 19). Similar results were given also by twelve plants which were grown on for more than 12 days four of them for 21 days. There was still no sign of differentiation of the cells of this sector. If the results of all these experiments are studied in Table 3 (p. 451), it can be seen by comparing the control apices with those in which Pj was removed, that the presence of the primordium is essential for the formation of the cambiform elements and desmogen strands, and that without it the merismatic cells of the stele below develop into parenchyma instead. Also by comparing rows 2 and 3 of the table, it can be seen that auxin does not cause the merismatic cells of the ring to differentiate into cambiform cells and desmogen strands, but does preserve them from differentiating into parenchyma. This preserving action may in the normal plant facilitate the differentiation into cambiform cells and desmogens under the infiuence of some other factor or factors. This other factor or group of factors is unknown, but clearly it is formed by the young primordium. The name desmin is suggested for it, and its effect will be called the desmin effect. Expt. 8. Pj removed Six further experiments were carried out in which P^ was removed. They all lasted 10 days, and they gave results similar to those of the P.^ removals of 6 to 9 days' duration. There was a gap seen in the stelar ring below the stump of the primordium, consisting of small parenchyma cells interspersed with scattered merismatic cells. In another six plants P^ was removed and its stump was auxinated; these experiments also lasted 10 days and gave results similar to those of the comparable series of Pj removals, showing a band of meristem cells below the stump of P^. In one of these apices, however, the primordium had been incompletely removed, as in Helm's experiments (1932); its base had grown slightly, and within it there was a very small strand of a few thick-walled conducting cells (Fig. 20). These experiments support the conclusions drawn from the removal of P,. Expt. 9. All visible primordia removed In twelve apices all the visible leaf primordia were removed, and the apices were then examined every 3 days and any subsequently formed primordia were removed also. The experiment lasted 21 days. In order to obtain twelve valid sets of results twenty-four apices were used, of which twelve died. In those which survived, transverse sections showed only a thin ring of merismatic cells, which was broken only in a few places by patches of parenchyma (Figs. 21, 22). Such a ring persisted down to levels at which in the control plants there was differentiation of phloem and xylem elements (Fig. 23). Still lower the merismatic ring of cells persisted, with no sign of leaf gaps nor of cambiform elements as judged by LjB ratios. Wardlaw (1950) working on Primula obtained rather similar results, but in his experiments there was a solenostele in the defoliated axis, with differentiated conducting cells, whereas m the present experiments only a ring of merismatic cells was found. The

13 The effects of leaf primordia on differentiation in the stem 457 Fig- 24 Fig. 20. Section of a Lupinus apex with Pj incompletely removed and its stump auxinated, after io days, showing a slight growth of the base of Pj and a small strand (S) of thick-walled cells within it. Figs. 21, 22. Sections of Lupinus apices with all primordia continually removed. Fig. 21 is a section 280^ below the summit showing a slight meduuation {Med.) by parenchyma. Fig. 22 is a section 570 ;i below the summit, at the insertion of an older leaf, showing a thin ring of merismatic cells (M.R.) with no cambiform elements. Both x 32. Fig. 23. A section through a normal apex, 280^1 below the summit for comparison with Fig. 21. The normal is much broader, x 32. Fig. 24. Section of a Lupinus apex with P3 removed, after 14 days, showing a parenchymatous gap in the stele with the cathodic trace of I^, marked kt. I^, in the gap.

14 458 B. S. YOUNG persistence of this ring was presumably due to auxin formed in and diffusing from the defoliated apex, which in all twelve plants was still functioning and wholly merismatic. In two of these apices a single new primordium had arisen in the 3 days after the last examination. It is remarkable that in these completely defoliated apices a thin merismatic ring persisted, although when only a single primordium was removed the sector of the ring' below it turned into parenchyma. The explanation may be that when only one primordium is removed, any slight merismatic growth which might otherwise have continued in the sector of the ring just below it is somehow inhibited by the development of the other sectors below the other young primordia. These results are consistent with the conclusion that the total leaf effect is to cause the differentiation of cambiform cells and their aggregation at certain points into desmogen strands. For they show that the stem apex alone can keep the cells of the meristem ring merismatic for a considerable time, probably by secreting auxin, but that in this species it does not cause cambiform cells nor desmogens to form. Why the merismatic ring occupies its actual position in the stem is unknown. The problem is similar to that posed by the normal axis, in which we do not know why the stele and the desmogens form where they do. Expt. 10. Pg removed There were twelve plants in this experiment, which lasted 14 days. In the sector of the stele below P3 there were no leaf traces leading out towards it, but instead there was a gap consisting almost entirely of parenchyma (Fig. 24). But through this gap ran the cathodic stipular trace of I^, which in the normal plant fuses with the median trace of the leaf corresponding with Pg. The median trace of Pg was just present at the time of operation, so that /^ needed only to make a trace connecting with that trace, of which the upper and outer part, leading out to Pg, must have de-differentiated. On the other hand when P2 is removed, its trace is not present. So the cathodic trace of /, cannot join with it, and instead it diverges to one side of the gap, which remains empty of traces. DISCUSSION The results reported show that in Lupinus albus there are at least three processes involved in the early differentiation of conducting tissues the retention of cells in a merismatic condition, their subsequent differentiation into cambiform cells, and their aggregation into desmogen strands. The causes of the differentiation of the desmogens and of other cambiform cells into xylem and phloem are a further question. As to the preservation of the merismatic state of the cells, the evidence is quite conclusive. The application of auxin to the stumps of removed primordia causes the cells in the sector of the stele below the stump to remain merismatic, at least so far as can be shown anatomically. These cells were at all times continuous with the merismatic cells of the apex during the periods of the experiments. So it should be stated not that auxin caused the formation of merismatic cells, but that it inhibited differentiation of merismatic cells to parenchyma. We may also conclude from the results that merismatic cells have a tendency, when uninfluenced otherwise, to differentiate into parenchyma. But the total effect of a leaf primordium leads to the differentiation of merismatic cells into cambiform cells, and thence to their aggregation into the desmogen strands. From this, therefore, together

15 The effects of leaf primordia on differentiation, in the stem 459 with the results of applying auxin to the stump of a removed primordium, one may infer that the primordium forms another factor or group of factors, to be called desmin. Support for these conclusions comes from the work of van Overbeek & others (1941), who cultured small embryos of Datura, and found that in order to obtain growth and differentiation they had to add an extract of coconut milk to the medium. Auxin alone led only to the increase in size of the embryo, but not to cellular differentiation: for without the coconut factor the embryos merely increased in size up to a maximum without further differentiation. The chemical nature of the active constituent of coconut milk is not known. Similarly, Skoog (1944) showed that auxin (indoleacetic acid) at concentrations of 0-2 parts per million prevented the differentiation of callus tissue in culture, but did not retard its growth in weight. We may therefore list the known and inferred causes of differentiation of conducting strands as follows: (1) The leaves determine their traces, at least in Lupinus albus. The traces are at first desmogen strands. (2) These desmogens are differentiated from a ring of merismatic cells which are kept merismatic by auxin. (3) Some other factor or group of factors, here called desmin, causes the formation of cambiform cells within the merismatic ring, and their aggregation at certain positions into desmogen strands. SUMMARY 1. Examination of the normal shoot apex of Lupinus albus shows that it consists at the extreme tip of a mass of merismatic cells, within which there is formed a ring of cells, delimited towards the centre by a parenchymatous medulla and peripherally by a parenchymatous cortex formed from the outer faces of the leaf buttresses. This ring of cells is wholly merismatic, and in it are formed the cambiform cells, aggregated into desmogen strands, which become the traces leading to the leaves. 2. Removal of the second youngest primordium led to the formation in the sector below it of a gap in the ring, composed of parenchyma. 3. When paste was applied to the stump of the removed primordium, the sector of the ring below the stump remained merismatic, and formed neither parenchyma nor cambiform cells. 4. It is concluded that auxin prevents the differentiation of the merismatic cells into parenchyma, but that their differentiation into cambiform cells and desmogen strands depends on some other factor, or group of factors, for which the name desmin is proposed. I wish to express my most grateful thanks to Mr R. Snow for his kindly help and criticism throughout the course of this investigation, and for help in presenting the results, and to Prof. T. G. B. Osborn for granting me the facilities of the botanical laboratories at Oxford.

16 460 B. S. YOUNG REFERENCES BALL, E. (1949). The shoot apex and normal plant of Lupinus albus. Amer. J. Bot. 37, 117. FLEET, VAN (1950). Cell forms and the common substance reactions etc. Bull. Torrey Bot. CI. 77, 340. GOODWIN, R. H. (1937)- The role of auxin in the leaf development of Solidago. Amer. J. Bot. 26, 43. HELM, J. (1931). Untersuchungen uber die Differenzierung der Sprossscheitelmeristeme etc. Planta, 15, 105. HELM, J. (1932). Uber die Beeinflussung der SprossgewebedifTerenzierung etc. Planta, 16, 607. LOUIS, J. (1935). L'ontogenese du systeme conducteur dans la pousse feuill^e etc. Cellule, 44, 87. OvERBEEK, J. VAN, CoNKLiN, M. E. & BLAKESLEE, A. F. (1941). Factors in coconut milk, etc. Science, 94, 350 SKOOG, F. (1944). Growth and organ formation in tobacco tissue cultures. Amer. J. Bot. 31, 19. SNOW, R. (1929). The young leaf as the inhibiting organ. New Phytol. 28, 345. SNOW, M. & SNOW, R. (1931). Experiments on phyllotaxis, i. Phil. Trans. B, 221, i. SNOW, M. & SNOW, R. (1947). On the determination of leaves. New Phytol. 46, 5. WARDLAW, C. W. (1944). Experimental and analytical studies of Pteridophytes, 4. Ann. Bot., Lond WARDLAW, C. W. (1947). Experimental investigations on the shoot apex of Dryopteris aristata. Phil. Trans. B, 232, 343. WARDLAW, C. W- (1949). Further experimental observations on the shoot apex of Dryopteris aristata Phil. Trans. B. 233, 415. WARDLAW, C. W. (1950). The comparative investigation of apices of vascular plants etc. Phil. Trans. B

17