Search

This experiment was carried out to make clear the influence of the height and diurnal range of temperature on the morphology of spikelets in normal rice plant, cultivars Akibare and Reimei, and on the type of proliferation in their mutant strains, both of which have been cultivated in green house or phytotron under natural daylength of summer and winter. It was observed on the normal strains that from about eight to twenty nine percent of spikelets botained in winter in green house and phytotron of low temperature developed their glumes; remarkable elongation and multi-venation of empty glume and lemma or differentiation of awn, contrasting the ones in summer having scarcely such develped one. On the mutant strain induced from cultivar Akibare, it was observed that in winter every spikelets in green house and phytotron developed the same kind of glumes as in normal strains, although in summer relatively fewer ones in green house did those glumes than in phytotron. As for the mutant strain of cultivar Reimei, it was clarified that in summer most part of spikelets in phytotron of low temperature differentiated pistil-type (P-type) proliferation in contrast with most of them in green house did spikelet-type (S-type) one, and in winter about twenty seven percent of spikelets in phytotron of low temperature did leafyshoot-type (L-type) one which was hardly observed in other experimental conditions of this study, besides most part of them in not only phytotron but also green house did P-type one. Considering that the proliferation progressed gradually to vegetative characteristics as the type changed from S-type to L-type going through P-type, results mentioned above suggested that small diurnal range of temperature particularly under lower one might promote more effectively the changing type of the proliferation into vegetative than larger diurnal one. Comparing those results with those of the effect of gibberellin and/or naphthalene acetic acid on the proliferation which had been reported previously by the authors, it was considered that proliferative changes induced by low temperature, its small diurnal range and short day seemed to have similar physiological back ground with that induced by those growth substances. Some discussions were undertaktaken on the fact that phenomena of spikelet proliferations in Gramineae had been observed mostly in Festucoid than Panicoid by many workers.

Water loss in hull and the effect of aerial humidity on the development of kernel were investigated. The results obtained are as follows: 1. The rate of water loss in the hull after detaching from panicle decreased, as ripening proceeded in both fertile and sterile grains. 2. The rate of water loss differed among sterile, fertile grains, and hull of fertile grains, being higher in the following order: sterile grain>hull of fertile grain>fertile grain. 3. The rate of water loss was much lower in normal than in malformed grain. 4. The water loss in panicle was affected by SiO2 application and it was higher in detached panicle in plots without SiO2 application. 5. By glume top clipping, the development of the kernel was inhibited very severely. But even though the top of glume was clipped, the development of kernel was not prevented so severely under high aerial moisture conditions. And even under the same moisture conditions, when the glume-top-clipped kernel was covered with a parafilm bag, the development of the kernel was better than that without covering of parafilm bag. From these results, it was shown that in the ripening of rice plant, the hulls play an important role in keeping high moisture condition in grain, thus enabling the kernel to achieve a normal development.

The effect of ultraviolet (uv) ray on rice (Oryza sativa L) kernel during its kernel development was investigated. Upper one-thirds portion of each flower glumes (palea and lemma) on the middle branches of panicles were cut off immeadiately after flowering. Panicles with the cut glumes were then covered with a vinyl film envelope of different light transmittance in uv range. Kernel development in the cut glumes was greatly inhibited by being exposed either to the natural light or to the light of uv range. Relative humidity, however, seemed not to have a great influence on the development of the glume-cut kernels. The inhibition was most outstanding when exposed to short wave uv rays, although it was lessened as exposed to longer wave uv rays. Inhibitory effect of visible light was very slight. Moreover, under dark conditions, kernels in cut glumes developed quite well beyond the untreated kernels and formed a Dahrma doll shape. In terms of kernel characters, uv radiation produced tapered kernels and promoted a browny coloring of kernel as well. Another experiment as to the degree of glume cutting revealed that under dark conditions, the effect of the extent of cutting on the inhibition was rather slight, although they were very conspicuous when exposed to the natural light. Rice flower glumes was found to intercept uv rays almost completely after the last stage of booting. These results indicate that exposure of young kernels to a short wave uv ray strikingly supresses their development and that rice flower glumes may take an important role in protecting their inside kernel through intercepting harmful uv rays. Rice kernels might substantially have an ability to develop beyond the volume of glumes though their development are usually restricted within the volume limit of their glumes.

In order to make clear the morphological nature of the 'compound 2-flowered' spikelet [Plate 1-1, 2] in Oryza sativa L., the present investigations were carried out from two aspects as follows; 1) Gross morphological ones of the acropetal orders of each bract (glume and lemma) on its own spikelet axis and those of the disposed direction of each bract relative to the primary (or secondary) rachis in the inflorescence. 2) Anatomical ones of the vascular courses of the 'compound 2-flowered' spikelet with two ovaries. As a result of the first investigation, a concept was composed, which is to be called 'isobract'. This concept is defined as a 'set' of the bracts among which each acropetal order is equal on spikelet axis and each disposed direction is equal or bisymmetric relative to primary (or secondary) rachis. For example, in Plate 1-8, there are five isobracts; (rg1, I), (rg2, II), (eg1, III), (eg2, IV), and (1, V) respectively. The palea (p) and the bract VI aren't isobract because of the fact that the former is the bracteole on floret akis. Hence it is considered that the development of the bract VI is caused by the prolification of spikelet primodium and not by the reversion of phylogenetically degenerated bract. As regards this problem, other authers' interpretations arc discussed, particularly those of Parodi (1941) and Schweickerdt et al (1956). In the other side, it became clear from the second investigation that two main vascular bundles of floret axis exist in the stalk of the 'compound 2-flowered' spikelet. One diverges from mother vascular bundle at the axil of the bract V and the other does in the same manner at the axil of the bract VI [Plate 2, vfa and vfa']. The mode of the vascular courses, where each main vascular bundle of floret axis diverges from that of spikelet one (mother vascular bundle), is plainly the same as that of the normal spikelet. These results demonstrate that the stalk of the deformed spikelet consists of three axes; one spikelet axis and two floret ones, the letters independently arise at each axil of the bract V and VI, and have grown together in each dorsal side. Therefore the 'compound 2-flowered' spikelet with two ovaries truly had two florets which has compounded ontogenetically. The hypothesis and Model II (Matuba, 1971) mentioned in previous report are thus confirmed.

Although the size and shape of the caryopsis in rice plants are determined by the capacity of the glume, the growth of the caryopsis and the floral glumes may be controlled by different genetic systems and thus they may have different developmental paths. Therefore, in varieties having potentially much longer caryopsis than the glume, the growth of the caryopsis may be affected by the glume leading to abnormal grain formation. One of such abnormal development of caryopsis is "notched" grain which is observcd with variable frequencies in some of the rice varieties. This investigation reveals an account of the mechanism of grain notching and the variation for notched grain frequency in rice. A total of 32 varieties and lines was grown in glasshouse as well as in field to investigate the varietal and environmental variation for the notched grain frequency and also the correlation of notched grain percentage with the unbalance between the length of the caryopsis and the glume. For allowing the caryopsis to grow freely, the top of the glume on one or two panicles in each of the plants was clipped off at two or three days after the anthesis. In the field, glume clipped panicles were protected by parchment paper bags. The length of glume-clipped caryopsis (T1) and untreated one (C1) were measured using ten caryopses randomly taken from each of the 6 plants in glasshouse and 20 plants in the field. T1/C1 ratio was calculated to indicate the degree of unbalance between the potential length of caryopsis (T1) and the longitudinal capacity of glume (C1). One hundred untreated caryopses from each of the plants were exarnined to determine the percentage of notched grain. The percentage was transformed into the degree of angle for the statistical analysis. Correlation between the T1/C1 ratio and the notched grain percentage was also analyzed. Varietal variation ranged from 0 to 90% for notcched grain and from 1.1 to 1.5 for T1/C1 ratio which indicated that the potential length of the caryopsis was 1.1 to 1.5times more than the longitudinal capacity of the glume. From these results, the materials used were classified into three groups: Group A, showing remarkably high, more than 80%, grain notching and high T1/C1 ratio in both the glasshouse and field conditions, group B, showing highly sensitive response of grain notching to the environmental conditions, and group C, showing mostly normal grains with low T1/C1 ratio. The glume-clipped caryopsis did not develop notched grain in all of the materials. Correlation between the glasshouse and field was high for C1 and T1, but it was rather low for notched grain percentage and thc T1/C1 ratio in some varital groups. Highly significant correlation between notched grain percentage and T1/C1 ratio (r=0.94) suggested that unbalance between the length of glume and caryopsis was the primary factor determinig the grain notching. From these studies, it may be concluded that, along with other factors, abnormal growth of the caryopsis restricted by the glume capacity results in notching of the caryopsis.

Grain size and shape are important traits in rice. However, it is not yet been fully understood how the grain size or shape is determined during plant development. In the present study, we analyzed the growth of rice grain in detail, using five cultivated strains (O. sativa) and one wide strain (O. rufipogon=perennis). The growth of rice grain was characterized by the initiation of rapid growth immediately after flowering and the abrupt cessation of the growth at maturing stage, suggesting a non-logistic growth. Also, relative growth rates in length and width and the allometric growth between length and width showed a non-logistic nature of growth. The final grain length correlated well with the maximal growth rate in length. But the final grain width was closely related to the growth period, not to the maximal growth rate. This shows that the grain size is controlled by different parameters, depending on the direction of the growth. In the cultivated strains, the minimal ratio of grain width to length, which was attained on the second to fourth day after flowering, correlated with the final ratio. Therefore, a final grain shape can be estimated by the minimal ratio. These results show that rice grain has a non-logistic and unique mode of growth not observed in the growth of glume or embryo.

Grain size is one of the important yield components and it has influences on grain yield as well as grain quality. However, it has not been well understood how grain size is determined. Certainly, glume size is one of the important factors affecting grain size. We analyzed in detail the growth of glume (floral glume) of several rice varieties differing in sizes of their glumes. 1. The growth of glume in both length (lemma length) and width (floret width) were well fitted to a logistic curve (Fig. 1). This indicates that the growth pattern of glume can be characterized by parameters estimated from the logistic curve. In every variety, the growth of the glume in length ceased earlier than that in width (Figs. 2 and 3). Thus, the growth of glume in length did not synchronize with that in width. The intervarietal differences of growth period did not correspond with those of the glume length. However, the maximum growth rate of glume in length correlated highly with the final glume length, indicating that the differences in the final glume length was due to the differences in the maximum growth rate of glume length among the variesies (Table 1, Fig. 2). on the other hand, in the glume width, we could not detect any distinct intervarietal differences in growth period or maximum growth rate (Table 2, Fig. 3). It must be noticed that the relatively wide glume of "Arborio" was caused by the prolonged growth but not by the high growth rate (Table 2). 2. The growth patterns of glume in length and in width are different and the length and width of glume may be controlled by different regulatory systems. In this point, we need to analyze the growth pattern of glume and the mechanism of determination of its size in more detail. 3. Intersecting lines comprising three straight lines could be fitted to the allometric growth between the glume length and width (Fig. 4). In every variety, the coefficient of allometry was minimum in the second section and maximum in the last (Table 3). Intervarietal differences were detected in the coefficients of allometry and the positions of intersecting points. It is clarified that the shape of glume is affected by both the ratio of growth rate in length and to that in width of glume and the time lag of the growth period in the width relative to that in the length. 4. The present results show that the size and shape of rice glume are controlled by various parameters and we cannot assign any single factor determining the size or shape of glume during its growing process.

The effect of nitrogen top-dressing (ntd) on the gibberellin (GA3, GA) - induced abnormality of spikelets was investigated during the two years, 1986 and 1987, with using the rice cultivar Sachikaze. Plants grown in 1/5, 000 a pots filled with paddy soil were top-dressed with different amounts of nitrogen at the maximum tillering stage, and one week after, the plants were treated with GA of 500 ppm until the final concentration was reduced to 50 ppm. Three levels of ntd with or without GA treatment were tested in six experimental plots; (1) No ntd (0 N), (2) Standard ntd (1 N), which was applied with 0.5 g of ammonium sulfate, (3) Three-times amont of ntd (3 N), which was applied with 1.5 g of ammonium sulfate, (4) GA application and 0 N (0 N+GA), (5) GA application and l N (1 N+GA), and (6) GA application and 3 N (3 N+GA). Plant height, and numbers of secondary branch and spikelet increased in all plots of ntd especially in those with GA apply. The percentage of spikelets with any abnormal organs in GA treated plants increased with ntd. A decrease in number of stamens to five or less, and pistil with three stigmas were prominent abnormalities observed in the study. Third stigma development was observed in 3 N plot even without GA application, showing highest 30% with GA application. The distribution pattern of abnormal organs in the panicle was similar in both additional glume and stamen, however, it was different in the pistil. Under high ntd level, GA application induced both suppression of stamen development and promotion of pistil differentiation under high frequency; GA supplied without ntd brought on increase in stamen development. These results indicate that the effect of GA in morphogensis of reproductive organs in rice changes with ntd conditions. GA was found to promote the effect of nitrogen in increasing female organogenesis, similar in effect to that of auxin applied exogenously.

The occurrence of unclosed glume grains and cracked grains in the paddy rice harvested with combine harvester was recognized by Dr. Suetsugu et al. in 1965. Such abnormal grains are regarded as a problem which should not be overlooked from the viewpoint of process of drying of high moisture-content paddy rice grains. However, such abnormal grains can also be found in the paddy rice before they have been harvested. In the present report, the authors deal with the differences in the occurrence of such abnormal drains according to rice varietal differences and according to rice growing conditions. The test results may be summarized as follows: 1) Unclosed glume grains occur frequently in the case of early-maturing varieties, while occur very seldom in the case of medium-or late-maturing varieties. However, even in the cases of varieties similar in early-maturing, it is clear that the occurrence of abnormal grains varies according to varietal differences. 2) Even in the case of early-maturing variety in which unclosed glume grains are liable to occur, if the heading date is adjusted so as to be delayed to the extent of heading date of medium-maturing variety, the occurrence of unclosed glume grains would be reduced to the extent similar to the case of medium-or late-maturing variety. 3) In the cases of direct-sowing rice plants and densely transplanted ones, there is an increasing trend in the occurrence of unclosed glume grains, as compared with ordinary transplanted ones. 4) Both in the cases of ordinary transplanting and direct-sowing, it shows increases in the occurrence of unclosed glume grains when rice seed is directly sown earlier on the paddy field or when rice seedlings are transplanted earlier. 5) In case where nitrogen is decreased in the amount of application, and phosphate is increased in the amount of application, comparing with the standard application amount, unclosed glume grains will occur more frequently. The same trend can be seen in case where potassium is decreased in the application amoun. Sincere acknowledgment is expressed to Dr. Suetsugu, Director of the Fukui Prefectural Agricultural Experiment Station, for the fact that the above tests have been achieved under his expert direction.

Vatious morphogenic abnormalities as shown in figures were observed in the rice spikelets which were grown under low temperature-day temperature 14-18°C and night temperature 14°C. Morphogenic abnormalities found in the spikelets may be summarized as followings: 1) changes of organs in number, especially increment in number, 2) remarkable phenomena of fusion of inter and/or intra organs of lodicules, stamens and ovaries, 3) appearance of organs of intermediate and bisexual structures, 4) changes in sex expression, promoting femaleness and depressing maleness, 5) continuance of vegetative growth in the spikelets with no formation of sexual organ, incepting only glume and branch primordia, 6) exposure of extraordinarily developed ovule or integument from ovary through the style canal, 7) frequent occurence of abnormally swelled tissue or organ, and 8) appearance of multi-floret spikelets, etc. From the results above mentioned it is considered that the morphogenic regulation in the meristematic tissue of spikelet primordia may become unstable and the disturbance of organ development tends to be occured, by the low temperature treatment. Some discussions were tried on the relationship between the appearance of morphogenic abnormalities due to a low temperature treatment and the level of some growth substances, auxins and gibberellin, in spikelet primordium, based upon the writers' experimental results previously reported.

On the way to comparative studies of the deformities caused by gibbefellin treatment in the spikelets of 17 rice varieties, the writer take the methodological position that the developmental studies of deformed spikelets are basically important in order to interpret scientiftcally the morphological nature of such spikelets, classify them systematically and to understand their developmental physiology experimentally. Since we have only fragmental knowledge on the development of deformed spikelets the constitution of them was analyzed theoretically and some hypotheses were drawn as follows. 1) The hypothesis on the development of normal spikelet : The apical meristem of spikelet primordium normally degrades right after the differentiation of a lemma primordium on it and then the other floral organs, these are, a palea, two lodicules, six stamens and a pistil, differentiate successively on the floral primordium initiated from the axil of the lemma primordium. This spikelet primordium generally develops in the normal spikelet that has the constitution as shown in model N (fig. 3-5.) The axis of the spikelet is not single but consists of two axes namely a spikelet axis and a floret one. The floret axis is a branch initiated from the axil of lemma on the spikelet axis. The palea corresponds to a prophyll on the floret axis in it's morphological nature and the two empty glumes do "sterile Iemma" on spikelet axis, two rudimentary glumes do the so-called "glume" on spikelet axis. Since the apical meristem degrades soon after the differentiation of a lemma primordium it is expected that the tip of main vascular bundle in spikelet axis exists in the region of the axis between the upper empty glume and the palea. 2) The hypothesis on the development of deformed spikelets : So far as the writer's observations in this study, the deformed spikelets are of two kinds. The spikelets of first group (Group I of the deformed spikelets) have the constitution (of a floret and those of second group (Group II of the deformed spikelets) have the constitution of bi-florets. a) On the development of the spikelets in group I. When the floral organs differentiate irregularly on the floret primordium under so me abnomal conditions, the spikelet primordium develops in a spikelet having variform floret. The typical variations observed in these spikelets are as follows. (i) The variations in the nerve number are from 1 to 6 in lemma, from 2 to 5 in palea (figs. 2-10, 14). (ii) The variations in the number of stamens are from to 0 and l0 (fig. 2-13). (iii) The variations in the number of pistil are from 0 to 3. There are various mediate types of multi-pistils. b) On the development of the spikelets in group II. When the apical meristem degrades after successive differentiation of two lemma primordiaon it, the spikelet primordium develops in a spikelet having bi-florets constiution because two floret primordia can be born on each axil of two lemma primordia. Though the spikelet primordium has a potentiality to develop two perfect florets in it (model II fig. 3-2), as differentiation of floral organs on each floret primordium takes place under various abnormal conditions in actual state, generally it can not have two perfect florets. Usually the lower florets are almost imperfect (figs. 2-6, 3, 4, 2, 7, 8) and the upper florets are perfect (fig. 2-4) or more (dominant than the lower ones (figs. 2-9, 12). Thus all of these spikelets are explained as modified types of the model II shown in fig.3-2 consistently.

1) This study was carried out to make clear the distribution of morphological abnor- mal spikelts induced by gibberellin. 2) It was indicated that the number of spikelets was increased remarkably with the treatment as previously reported and the increase was more conspicuous in upper primary branches than in lower ones. 3) Main morphological abnormalities, such as inception of additional glumes, chandes of stamens and pistils in number were found more frequently in the upper portion of the primary branches. 4) Judging from the above results, it is considered the morphogenic effects of gibberellin on panicle development are more remarkable in the upper or distal portion of panicles or primary branches. This tendency also supported by the writers' studies on the effects of gibberellin on the vascular development in panicles. 5) On the other hand, so called "white hulls" were found frequently in the basal portion of lower primary branches. So, it is concluded that the initiation of "white hulls" should be brought about by a quite different pathway from that of the morphogenic abnormalities. 6) The difference in frequency of the abnormality in glume, stamen and pistil was indicated and discussed from the stand-point of the ontogenic changes in form and related morphogenic substances in the meristematic tissue of spikelet primordium.

Some morphological observations were carried out on the proliferations in the mutant spikelets induced by X-ray irradiation in Oryza sativa L. var. Reimei. X-ray was irradiated on seeds at the dosage of 20 kR. in 1966, followed by sowing in soil. Spikelets which were used for observation had been sampled in the year from 1968 to 1972. Results which have been obtained were as follows; 1) One or more spikelet-like organs were found in the spikelet. 2) One or more leafy-shoots which have been composed of many leaf-like glumes were found in the spilelet, the tips of which had (a) one or more spikelet-like organs, (b) incomplete ovary, (c) stigma-like intermediate organs having no ovarian structure, or (d) vegetative shoot having no reproductive organ. There were general tendencies in those spikelets having proliferations mentioned above, such as (1) rudimentary glume had no change, (2) lemma and palea had many veins in themselves exhibiting slender forms, (3) the position of spikelet-like organs to differentiate was above that of palea where reproductive organs differentiated usually, and no reproductive organ was found in that place. It was considered that these spikelet-like organs or leafy shoots were the typical proliferations, and those proliferations would have been induced just before the time of differentiation of reproductive organs on the apical meristem of spikelet primordium, and that those proliferations might be initiated as the result of the supression of transformation of apical meristem from vegetative stage to reproductive stage in a narrow sense. Judging from the results and considerations mentioned above, it was presumed that the apical meristem of the spikelet primordium which was going to differentiate proliferation maintained relatively larger size and more differentiated organization than those of the control at the time when stamen primordia differentiated. Some discussions were undertaken on the physiological and genetical background of the development of the proliferations.

In the previous paper, a hypothesis (Model N), interpreting the morphological nature of the normal spikelet in rice plant, was proposed. And it was postulated that the main vascular bundle in the axis of the normal spikelet consists of two kinds; that of spikelet axis (rachilla) and that of floret one, and their connective point presents in the axis between upper empty glume and palea. In this paper, the course of vascular bundle of the normal spikelet is traced with serial-microtome sections. The main vascular bundle extends to lateral direction in the level just above the abcission layer (Plate 1-1∼3) and forms a ring (Plate 2-1), that is, a vascular cylinder. As shown in Plate 1, it is clear at higher level that the main vascular bundle has a large gap (arrow symbols in Plate 1) between the both bases of midrib trace of lemma and palea (a and o in Plate 1). And the main vascular bundle forms a solid column from the gap to the base of pistil. To research the detail of the vascular course around the gap is an important problem. The trace of midrib of lemma (a in Plate 2-2) detaches itself from the vascular cylinder at the ventral side, forming an incomplete gap on this cylinder (Plate 2-1∼3). The half tissue of the vascular cylinder containing left and right parts of the leaf trace "a", increases in the thickness at the ventral side and two lateral leaf traces diverge into the latero-ventral position from the both parts (b and c in Plate 2-3). These traces connect with each lateral veins of lemma. Subsequently the both parts of the vascular cylinder join together inside of the cylinder and synchronously detach themselves from the dorsal arc of the cylinder (Plate 2-4∼5). In this way, the main vascular bundle of solid column type is composed at the ventral side, namely at the axil of lemma (vfa in Plate 2-6, 7). And this main vascular bundle diverges the traces into the flower organs; palea, lodicules, stamens and pistil (see Plate 1). However, it does not always occur in palea as follows. The dorsal arc of the vascular cylinder divides into four fragments after the divergence of a trace (c2 in Plate 2-5, 7), forming a gap, into upper empty glume. Two fragments of outer side (m and n) supply a trace to each marginal veins of lemma. Two remnants (w and z) join again each other as shown in Plate 2-8, and connect with the midrib of palea (o in Plate 2-11). Therefore, the midrib of palea, not producing any gaps, is classified into two kinds according to its origin (one; "o" in Plate 1, the other; "o" in Plate 2). The main vascular connection between "cylinder" and "solid column" is observed at the nordal region, where axillary axis (floret axis) branches out from main axis (spikelet axis), as shown in other grass spikelets, for example, Triticum and Bromus. The vascular system between the main vascular bundle of solid column type and traces of the flower organs resembles to that of floret in other grass spikelets, except that the traces of palea does not always diverges from the main vascular bundle of solid column type. Moreover it seems that these vascular courses in the normal spikelet are quite alike an unilacunar type as usually seen in many dicots. The conclusion obtained from these comparative anatomical aspects is that the main vascular bundle of solid column type at axil of lemma is considered to be a branch trace in connection with the vascular cylinder. In other words, the vascular cylinder and its branch trace are correspond to the main vascular bundle of spikelet axis and that of floret one, respectively. It can be said that our conclusion does not contradict with the hypothesis (Matuba, 1971) shown previously. Such anatomical analysis leads us to a conclusion that the axis in the normal spikelet consists of two kinds; spikelet axis and floret one, in spite of the external appearance as same as single axis. [the rest omitted]

1. In a floret of barley there are three anthers. 0ne of them is located at the side of outer glume while the other two are located at the side of inner glume (fig. I). The former is herewith called the first anther and the latter are called the second and the third anther respectively. 2. The first anther is characterized by the angle made between the lower ends of one-side locules and opposite-side ones at the attaching point of anther filament. This angle is called by the auther an anther's angle. The first anther's angle is, in many cases, remarkably greater than those of other two (fig. II). Meantime, no difference is recognized between the anther's angle of the other two. 3. The first anther is not only peculiar in the above-mentioned morphologycal point, but is also functionally inferior to the other two, showing the less diameter of pollen grains as well as delayed and smaller splitting of anther. 4. The first anther's angle in each bred line of a variety "Moravia" shows a positive significant correlation with the percentage of sterility of floret in them. Still more, when the ear is divided into three parts of upper, middle and lower, the angles of first anther in the upper part is greater than those in the remained two parts, in correspondence with the occurrence of sterility of floret that is also higher in the upper part than in the others. Within a plant, the first anther of the longest culm, that is usually consisted of main culm, is provided with a smaller angle than those of the shorter culm. 5. In view of those points, it is suggested that the first anther's angle might have a close correlation with the sterility of floret directly or indirectly.

Interaction between a progress of the kernel senescence and the transition in dehydrogenase activity in the spikelet tissue has been examined in a variety of rice plant (c. v., Koshijiwase). It was found in the rachilla that a disappearance of the enzyme activity with TTC reaction is the earlist at a part deserving of absciss layer in the pedicel of tissues, i. e., rachilla, husk, embryo, and endosperm. The proper and natural rout becames senile though the rachilla, empty glume, husk, vascular bundle of endosperm before occurring endosperm senescence in the spikelet tissue. Such a fact was found by a histo-chemical investigation. Therefore, there seems to be that to protect a senescence of the rachilla supporting a spikelet means to perfect the maturity.

In order to determine the glume hair density, number of hairs located along outer glume edge were counted under an anatomical biocular microscope, for five grains in each variety, about 1251 cultivated rice varieties being collected from the world wide rice growing countries. Wide variations were observed in glume hair density, and the authors, therefore, classified them arbitrally into 7 classes according to the mean number of hairs, namely, none (below 5) rare (5-35) very small (35-50) small (50-60) medium (60-70) many (70-80) and heavy (over 80). The results were compiled in table 1. From this table we can see that in northern districts such as Japan, Korea, Manchuria and North China, U.S.S.R. and U.S.A., the modes appeared at the medium class, and some heavy class types were also found among them. On the contrary, in southern districts such as Central and South China, Formosa, India and Java, the modes lowered themselves to the small class and no heavy class varieties were found. A peculiar variation was found in the Phillippine varieties where most belonged to none or rare classes. Such glaberima types were also exceptionally found in the Japanese teratological varieties. It is generally concluded that the glume hair is denser and longer in the varieties which came from the northern countries than in those from southern countries.

Barley's sterility "Schartigkeit" was recogmzed in the south-west district of Japan, especially in all prefectures in Kyushu in 1946. The sterility was noted in various varieties of barley, in partial or general on their head, especially in "Golden melon" most remarkably. At the farm of the Miyazaki Agricultural Experiment Station, "Golden melon" and others were sterilized perfectly and did not obtained any grain. This phenomena were regarded so important upon the agronomic practices and breeding experiments, that some investigations were performed at the farms of Kyushu Branch of Agricultural Experiment Station and of the all prefectures in Kyushu and others. The summary of the investigations is as follows: (1) The most unusual conditions in the climate during the barley growth period are that the temperature rised suddenly from the end of January on wards, and the amount of percipitation was great from the end of Februaly to the early day of May. (2) The culm was very shorter in all varieties of wheat and barley in the year than in the ordinary ones. Especially the varieties or the culms, in which plentiful sterile florets were recogoized did not ear, and these plants resembled to the "Sitzenbleiben" in late seeded barley of winter type. (3) The glume of the sterile floret was normal in form, but the development of male and female organs were somewhat imperfect. And especially the forms of the anther and of the pollen grain were abnormal, but malformed or organless plant was not found. (4) The tendency of cleistogamy was very remarkable in the year, and it is noticeable that on some varieties or culms late-eared, the fertilization was completed several days earlier than the earing date. (5) The distinction of this sterility may be seen easier on the two-rowed barley than on the others, but no relation was recognized between the forms of ear (two, four, or six rowed) and the such sterility. (6) Such sterility was found abundantly in the winter type varieties and less in the spring types. And in the same variety, sterility were found less in the earlier seeded plants than in the later seeded ones. (7) The correlation between the sterility and the date of earing has not been recognized, but the sterility was somewhat more notable in the later earing plants than in the earlier earing ones. (8) On the same variety, the sterility was found more remarkably in the warm region than in the cold one in the south-western Japan, viz., the warmer condition is the profitable for the outbreak of the phenomena. (9) Negative correlation was recognized between the sterilty and the earing degree, viz. the sterility is of high per cent in such plants, as the internode under the ear does not elongate so as the ear appears upon the sheath of the uppermost leaf. (10) The sterility was found not only in barley but also in rye and oats, and less in oats but abundant in rye. The sterility was also found in the rye grown in the glasshouse as about the same degree as in the open field under natural conditions.

1. Two varieties, Taihei-mochi as a variety of "many-panicle" and Toyokuni as that of "heavy-panicle" type were subjected to an experiment, in which later tillers of plants other than the limited number of stems, ie. 1, 5, 10 or 15 including main stem, were cut off as soon as they appeared, and the plants were compared with those left free. 2. Compensation effect of tiller limiting was most remarkably observed in increase of average panicle weight, the increase being mainly caused by increased average number of grains. Number of panicle branches, panicle length, and proportion of fertile grains to the total number of grains increased in this order. Increases were the cases also in number of nodes and length of main culm. 3. The conspicuous increase of the number of grains on single panicle as an effect of tiller limiting is due largely to intensified density of glume setting. 4. With regard to the compensational effects, ear growth and enrichment predominate over stem elongation; and elongation of uppermost internode does over that of the other part of the stem. 5. Another phase of compensation effects was acceleration in development of main-stem leaves, elongation of plant hight and heading time. Promotion of the ear growth and enrichment, at that time, predominates also over acceleration of the growth of other vegetative organs. 6. The grades of increase or acceleration by tiller limiting were larger in Taihei- mochi than in Toyokuni as to number of main-culm nodes, heading time, main-culm length and perdentage fertile grains. Acceleration of stem elongation became markedly to decline towards heading time, and the declination was more distinctive in Toyokuni than in Taihei-mochi. In number of panicle branches, somewhat larger grade of increase was recognized in Toyokuni, while no appreciable difference was shown between the two varieties as to increased grades in panicle length, single-panicle weight and number of grains of single panicle. Possible mechanisms of these phenomena were discussed.

In 1955, experiments were carried out to know how pollen germination may be affected by low temperature given at different stages of primordial panicle development, comparing with non-treated control. Pot cultured plants of variety "Norin-22" were exposed to low temperature of 17°C at 6 succeeding different stages of 6 days period during primordial panicle development. Fresh pollen taken from flowers of each treated plots were subjected to germination tests on starch paste medium. Results were summarized in Figure 1. According to the figure, we can find an appreciable reduction in pollen germination in the plot treated by low temperatures during the heading and the blooming stage (Plot F), compared with those in other plots, the differences being statistically significant at the one percent level. The plot treated during the reduction division stage (Plot D) showed also lower percentage of germination than other plots except Plot F, and these differences were proved to be significant at five percent level. While among other plots (Plot A, B, C, E and control) no difference was found in the germination percentage. The percentage of sterile glume was also measured with the same material plants in all plots. The percentage of sterile glume in Plot F was observed to be higher than those in other plots. Those in Plot D and E were next to that in Plot F, and most of the sterile glume in Plot D belonged to so called "Hakufu", whch means dead white coloured glume. Other plots except Plot D, E and F showed very few differences compared with the control.