Search

Premise of the study Root border cells are programmed to separate from the root cap as it penetrates the soil environment, where the cells actively secrete >100 extracellular proteins into the surrounding mucilage. The detached cells function in defense of the root tip by an extracellular trapping process that also requires DNA, as in mammalian white blood cells. Trapping in animals and plants is reversed by treatment with DNase, which results in increased infection. The goal of this study was to evaluate the role of DNA in the structural integrity of extracellular structures released as border cells disperse from the root tip upon contact with water. Methods DNA stains including crystal violet, toluidine blue, Hoechst 33342, DAPI, and SYTOX green were added to root tips to visualize the extracellular mucilage as it absorbed water and border cell populations dispersed. DNase I was used to assess structural changes occurring when extracellular DNA was degraded. Key results Complex masses associated with living border cells were immediately evident in response to each stain, including those that are specific for DNA. Treating with DNase I dramatically altered the appearance of the extracellular structures and their association with border cells. No extracellular DNA was found in association with border cells killed by freezing or high-speed centrifugation. This observation is consistent with the hypothesis that, as with border cell extracellular proteins, DNA is secreted by living cells. Conclusion DNA is an integral component of border cell extracellular traps.

We describe the unique development and branching of lobed thalli in Hydrobryum japonicum. Lobe formation begins with meristem initiation at random sites near the thallus margin fringed by protective tissues. As the protective tissues are successively peeled off particularly in the growing new lobes, the lobes become naked and then become fringed again by new protective tissues that develop from the marginal part of the new meristems. Subsequently the meristems become less active and are differentiated into parenchymatous ground tissue at maturity. The random pattern of meristem formation during the sporadic development gives rise to a nonorderly branching pattern of the thalli. Some other lobes (∼10%) are regenerated from injured parts of the thalli. The vegetative shoots arise endogenously near the thallus margin and are enclosed by the nonvascular strand nets. The rudimentary shoot apices remain embedded in the thalli. The thalli, though remarkably different from typical roots of other angiosperms, might be extremely transformed roots.

A spontaneous mutant, Abs2, that does not abscise any organs despite an apparently normal pattern of growth and senescence was isolated from among plants of Lupinus angustifolius cv. ‘Danja’. Abs 2 was found to be a recessive single gene mutation, and it was proposed that the gene for the original mutant phenotype, referred to as Abs2, be designated abs1. An artificially induced mutant allelic to abs1 was also obtained and a non-allelic mutant phenotype, Delabs (delayed abscission), which was designated abs2. Morphological and cytological features of the abscission process under conditions of natural and ethylene-induced senescence were compared in the wild-type parent and Abs2 mutant. In the parent genotype abscission under natural conditions is similar to many other species, consisting of a stage of cell division forming an abscission zone, activation of the cytoplasm of zone cells, dissolution of the middle lamella, disorganization of fibrillar wall structure, and cell separation. A slightly different pattern of abscission zone development was observed for ethylene-treated explants of the parent, mainly with respect to features of cell division and cell enlargement. In Abs2 no abscission occurred for any abscission sites under conditions of natural senescence or with ethylene treatment of small shoot explants. However, relatively normal abscission zones were differentiated at all sites in the mutant except that extensive cell wall disorganization did not occur. Ethylene production by leaves or other organs of the mutant was no different from that of Danja. Application of copper salts or hydrogen peroxide, droughting, waterlogging, or application of abscisic acid (ABA) increased ethylene production equally in both genotypes but did not result in abscission in the mutant. Release of root cap border cells, the only other cell separation process examined, was similar in each genotype. The study concludes that the mutation is quite specific to the abscission process and may be due to a lack of or delay in the expression of hydrolytic enzyme(s) associated specifically with abscission zone differentiation and separation.

The ultrastructure of root cap columella cells was studied by morphometric analysis in wild-type, a reduced-starch mutant, and a starchless mutant of Arabidopsis grown in microgravity (F-microgravity) and compared to ground 1g (G-1g) and flight 1g (F-1g) controls. Seedlings of the wild-type and reduced-starch mutant that developed during an experiment on the Space Shuttle (both the F-microgravity samples and the F-lg control) exhibited a decreased starch content in comparison to the G-1g control. These results suggest that some factor associated with spaceflight (and not microgravity per se) affects starch metabolism. Elevated levels of ethylene were found during the experiments on the Space Shuttle, and analysis of ground controls with added ethylene demonstrated that this gas was responsible for decreased starch levels in the columella cells. This is the first study to use an on-board centrifuge as a control when quantifying starch in spaceflight-grown plants. Furthermore, our results show that ethylene levels must be carefully considered and controlled when designing experiments with plants for the International Space Station.

While water shortage remains the single-most important factor influencing world agriculture, there are very few studies on how plants grow in response to water potential, i.e., hydrotropism. Terrestrial plant roots dwell in the soil, and their ability to grow and explore underground requires many sensors for stimuli such as gravity, humidity gradients, light, mechanical stimulations, temperature, and oxygen. To date, extremely limited information is available on the components of such sensors; however, all of these stimuli are sensed in the root cap. Directional growth of roots is controlled by gravity, which is fixed in direction and intensity. However, other environmental factors, such as water potential gradients, which fluctuate in time, space, direction, and intensity, can act as a signal for modifying the direction of root growth accordingly. Hydrotropism may help roots to obtain water from the soil and at the same time may participate in the establishment of the root system. Current genetic analysis of hydrotropism in Arabidopsis has offered new players, mainly AHR1, NHR1, MIZ1, and MIZ2, which seem to modulate how root caps sense and choose to respond hydrotropically as opposed to other tropic responses. Here we review the mechanism(s) by which these genes and the plant hormones abscisic acid and cytokinins coordinate hydrotropism to counteract the tropic responses to gravitational field, light or touch stimuli. The biological consequence of hydrotropism is also discussed in relation to water stress avoidance.

In the unusual aquatic Podostemaceae, the root is the leading organ of the plant body and is variously compressed and submerged as it adheres to rock surfaces in rapid water. In an anatomical comparison of the root apical meristems and root caps of 33 species that represent the major lineages of the family, the dorsiventrality of root meristems varied and was classified into four patterns: (1) The root cap is produced outward from a nearly radially symmetrical meristem. (2) The meristem and root cap are markedly dorsiventral; the outermost cells of the hood-shaped cap are acroscopic derivatives from bifacial initials on the ventral side, while the pattern on the dorsal side is similar to pattern 1. (3) Bifacial initials are on both the dorsal and ventral sides. (4) No root cap is present. An evolutionary polarity may be evident from pattern 1 to 2 and then to 3. Pattern 2 arose in the early evolution of the subfamily Podostemoideae and subsequently, pattern 3 arose in species with crustose roots, while the least specialized pattern 1 is retained in Tristichoideae and Weddellinoideae. Pattern 4 characterized by caplessness may have appeared recurrently in Tristichoideae and Podostemoideae. These evolutionary changes in the meristem preceded the specialization of external root morphologies.

The root apical meristem of Equisetum diffusum Don has a prominent four-sided pyramidal apical cell with its base (distal face) in contact with the root cap. Derivatives (merophytes) that contribute to the main body of the root are produced from the three proximal faces of the apical cell. The first division of a proximal merophyte is periclinal to the root surface separating a small inner cell from a larger outer cell. The inner cell is the precursor of the vascular cylinder. The larger outer cell is the precursor of the epidermis, cortex, endodermis, and pericycle. Radial sectors, established early in the development of the cortex, alternate with sectors in the vascular cylinder. These developmental steps show quite clearly that early root development in Equisetum is markedly different from that of most ferns.

Root tips of Pisum sativum cv Alaska were excised at seven excision distances ranging from 250 to 5,000 ,um from the root cap/body junction. Three response types resulted: 1) Type AOne to three regenerated apices grew in the same plane as the root axis; 2) Type B-One to three regenerates emerged either longitudinally from a swollen wound site or at angles >900 from the root longitudinal axis; 3) Type C-One to three lateral roots emerged at 900 from the root axis. The response type was dependent on the excision distance: Type A response resulted after excisions of 250 and 500 ,m, Type B after excisions of 1,000 and 1,500 ,m, and Type C after excisions of > 1,500 ,m. The vascular patterns of the regenerated roots varied with excision distance. Most regenerates resulting after being excised at 250 Mm exhibited variations in xylem strand number. Most of those resulting after being excised at 500 Mm were "polystelic" and few of these showed variation in xylem strand number. With increasing excision distance the number of regenerates with variation in xylem stand number or with a "polystelic" condition decreased. Both the type of regeneration response and the vascular patterns within the regenerates correlated with the state of differentiation of cells at each excision distance.

Primary roots of Zea mays cv. Yellow Dent growing in an electric field curve towards the anode. Roots treated with EDTA and growing in electric field do not curve. When root cap mucilage is applied asymmetrically to tips of vertically-oriented roots, the roots curve toward the mucilage. Roots treated with EDTA curve toward the side receiving mucilage and toward blocks containing 10 mM CaCl2, but not toward "empty" agar blocks or the cut surfaces of severed root tips. These results suggest that 1) free calcium (Ca) is necessary for root electrotropism, 2) mucilage contains effector(s) that induce gravitropiclike curvature, and 3) mucilage can replace gravitropic effectors chelated by EDTA. These results are consistent with the hypothesis that the downward movement of gravitropic effectors to the lower sides of tips of horizontally-oriented roots occurs at least partially in the apoplast.

Rhodamine-phalloidin was used to determine the distribution of actin microfilament bundles (mfb) in cells thought to be the site of gravity perception (statocytes) in coleoptiles and root caps of Zea mays and Hordeum vulgare. In coleoptile cells, amyloplasts were usually observed in close proximity to thick mfb, which often appeared to divide into finer mfb adjacent to individual amyloplasts. The nucleus in these cells was surrounded by an extensive network of mfb, which were connected to thicker transvacuolar mfb. Columella cells of the root cap contained an extensive reticulum of fine mfb throughout the protoplast, but lacked the much thicker mfb seen in coleoptile cells. The distribution and extent of mfb observed in fixed cells correlates with patterns of streaming and amyloplast movement seen in living cells. A possible role for actin mfb in the perception of gravity is discussed.

A B S T R A C T Using 3-day-old seedling roots of Zea mays L., cv. Kelvedon 33, it was possible to remove the root cap by a simple surgical manipulation without damage to the root proper. By a further small cut, the quiescent center (QC) itself was isolated. This double-convex lens-shaped tissue piece 100 X 250,um is composed of 1000-1500 cells representing only 0.25 mm3 in volume. The explant was demonstrated unequivocally by 3H-thymidine incorporation before excision and then by autoradiography to be composed of the specific cells usually designated the quiescent center. Using sterile techniques, the QC's were placed on nutrient agar slants and allowed to grow in culture. Of a number of nutrient media tested, only a medium supplemented with organic nitrogen components, ind6leacetic acid, kinetin and inorganic nutrients plus sucrose (S2M + K -2,4-D) was effective in eliciting development. Thirty to 40 percent of the 150 isolated QC's grown on this medium formed elongated roots, up to 2 cm in length in 3-4 weeks. Roots developing on agar medium showed in their proximal portion a vascular pattern with 5-6 metaxylem elements or variations of this pattern, but as the root elongated, the vascular pattern was progressively reduced in complexity at the more distal end to a small central group of metaxylem elements. When agar-grown roots were transferred after one week in culture to a liquid nutrient medium of the same composition, the initially reduced vascular pattern evident in the proximal tissues became progressively more complex in the distal portion of the root and after 2 cm of elongation, showed an essentially normal primary vascular tissue pattern characteristic of the seedling root.

A basic protein, lysozyme, labeled with fluorescein isothiocyanate, is readily taken up by roots of lasione montana. Most of the protein taken up is tightly bound to cell walls of the roots. Fluorescent protein is diluted in the growing region of a root as cells elongate and divide. Fluorescence remains in mature nongrowing regions and root cap cells for one to two weeks. Redistribution and translocation of the protein within the root is minimal or nil. A layer of chloroform-soluble material that prevents lysozyme from interacting with stem cell walls exposed to fluorescent lysozyme was found on stems of germinating lasione montana. SOME RHIZOSPHERE MICROORGANISMS secrete exoenzymes that can induce pathological conditions in plant roots (Jones, 1909; Ulrich and McLaren, 1971). Not all proteins taken up by roots are enzymes, and nonenzymatic proteins can influence root growth and cell function. Lysozyme (as a protein; there is no known substrate for this enzyme in roots) has been found to inhibit root growth and has minimal nutritional value as a nitrogen source (Ulrich, Luse, and McLaren, 1964). Uptake of basic proteins by plant roots has been studied with barley, onion, tomato, and vetch (Seear, Bradfute, and McLaren, 1968). Penetration is rapid and the amount of lysozyme taken up can exceed 1 mg per g of fresh barley roots. About 90 % of the protein is simply adsorbed by ion exchange in root cell walls (McLaren, Jensen, and Jacobson, 1960). Location of the remaining 10 % of the lysozyme taken up is less precisely known; by means of autoradiography, radioactive lysozyme has been found in cytoplasm and nuclei of cells several layers beneath the epidermis (Jensen and McLaren, 1960). Translocation of lysozyme to leaves has also been observed (Ulrich and McLaren, 1965). With fluorescence microscopy, fluorescent proteins have been found in cytoplasm of damaged cells (Drew, Seear, and McLaren, 1970). By working with a plant with thin roots (only 14 cells across in the zone of elongation), the pattern of lysozyme uptake in living roots can be examined under a microscope without resorting to fixation and sectioning. Thus the time sequence of events at any interval of an experiment can be followed. This study concerns the uptake of lysozyme, labeled with fluorescein isothiocyanate, by roots of lasione montana (Burstr6m and Gabrielsson, 1964) and what becomes of the Received for publication 21 May 1973. 2 Present address: Department of Biology, M. I. T., Cambridge, Mass. 02139. protein as roots continue to grow. Some observations on the lack of penetration of labeled lysozyme into shoots have been interpreted in terms of shoot anatomy. MATERIALS AND METHODS-Seeds of lasione montanta L., Campanulaceae, (a gift from Professor H. Burstr6m), were surface sterilized with 1 % sodium hypochlorite in soapy water for 30 min on a shaker. The sterile seeds were then washed with sterile distilled water 4-5 times before being placed on an agar in disposable box type petri dishes, 50 x 12 mm. Each dish was set on edge by taping against a vertical wall so that germinating radicals would grow downward along the agar surface rather than into the agar. The agar, 1 %, contained Ca(NO3)2 and KH2PO4, each at 1/5000 M. Seedlings were allowed to germinate for 50 hr in the dark at 78 F. All solutions and glassware were previously autoclaved to minimize contamination by microorganisms. Twelve seedlings with approximately equal lengths (about 1 mm) were picked from the germinating plate and lined in a row upon a new agar plate. The growth rate of each root was measured with the aid of a microscope. Plants with roots growing at the same rate were chosen for experiments and only normal plants were used. All treatments were performed with an individual seedling suspended in a drop of test solution on a slide. The slide was placed in a petri dish containing two layers of wet filter paper to keep the atmosphere moistened. After the treatment, seedlings were rinsed with control solution and returned to the standing petri dish. Root growth was followed during and after the treatment. The control solution consisted of Ca(NO3)2 and KH2PO4, both 1/5000 M, at pH 7.2. Test solutions contained in addition egg white lysozyme or a fluorescent lysozyme (Fl-lys).

A B S T R A C T Zea mays L. root development from the coleoptilar node was observed by light and electron microscopy. Roots developed opposite collateral vascular bundles in the coleoptilar nodal region. Three distinct histogens (stelar, cortical-protoderm, and root cap) became evident in early development. In median sections of the young roots, root cap and cortical regions formed a "hat" configuration over the stelar region. As the root matured, this "hat" developed centripetally to encapsulate the stelar region. Central core cells of the root cap were characterized by having numerous dictyosomes, amyloplasts, vacuoles, and thin cell walls. As these cells matured into outer or peripheral cap cells, the Golgi vesicles became hypertrophied. These hypertrophied vesicles contained a granular PAS-positive material which accumulated between the plasma membrane and the cell wall and formed a thick layer. As the PAS-positive material passed through the cell wall, it changed to a fibrillar texture. A PAS-positive material similar to that in the outer root cap cells was found adjacent to the outer walls of the protodermal cells. In median sections, PAS-positive material was not present in the promeristem region. Root cap cells as well as parent cortical cells were crushed as the young root forced its way through the parent tissue.

Graviperception by plant roots is believed to occur via the sedimentation of amyloplasts in columella cells of the root cap. This physical stimulus results in an accumulation of calcium on the lower side of the cap, which in turn induces gravicurvature. In this paper we present a model for root gravitropism integrating gravity-induced changes in electrical potential, cytochemical localization of calcium in cells of gravistimulated roots, and the interdependence of calcium and auxin movement. Key features of the model are that 1) gravity-induced redistribution of calcium is an early event in the transduction mechanism, and 2) apoplastic movement of calcium through the root-cap mucilage may be an important component of the pathway for calcium movement.

Nuclear DNA of meristematic, epidermal and root cap cells from the roots of three vascular plants-the cryptogam, Equisetum hyemale L, and the phanerogams, Tradescantia Clone 02 and Hordeum vulgare L.-was measured with quantitative Feulgen microspectrophotometry. Epidermal cells of all three species and root cap cells in both phanerogams contained up to 8fold the amount of nuclear DNA found in their respective meristematic telophase nuclei. In general, the large amounts of nuclear DNA parallel development and differentiation in the epidermis regardless of phylogeny, habitat, or degree of domestication. However, comparisons of the increase in nuclear DNA contents in the various epidermal cell types among these three species suggest that the mechanisms giving rise to these increases may differ phylogenetically and may represent another character in which cryptogams and phanerogams diverged in their evolution. NUCLEAR DNA LEVELS of epidermal cells of E. hyemale increase by strict doublings of the genome. This distribution suggests that endoreduplication is the mechanism responsible for increased DNA content. The relatively unspecialized root cap cells of E. hyemale show no increase in nuclear DNA content, but root cap cells of the phanerogams increase their nuclear DNA at least twofold. The epidermal cells of both Tradescantia and H. vulgare show a nondiscrete distribution of DNA contents, above and quite distinct from the doubling pattern seen in E. hyemale. We believe that gene amplification (or differential replication of the ge

The developmental anatomy of Zamia pumila L. root apices was studied during light-induced nodulation. Dark-grown roots had an apical organization identical to that of other cycads and similar to that of other gymnosperms. A distinct protoderm was not observed in these roots, which had a large open meristem and a root cap with a well-defined columella. During nodulation the meristem became reduced in size, and its constituent cells became vacuolate until all but a few resembled ground tissue. The root cap senesced during nodulation, and a recognizable root cap was absent from mature nodules. A file of densely cytoplasmic cells with centrally positioned nuclei developed in the nodule cortex. This layer was continuous across the nodule apex and was identical to the presumptive algal-zone described previously by other authors. Light-induced nodules branched dichotomously and were identical to algal-free nodules described by other authors. In dichotomously branched nodules, each lobe was covered by a parenchymatous mantle analogous to a root cap. A unicellular layer similar to the presumptive algal zone spanned the gap between opposite nodule lobes, and extended beneath each lobe before terminating in the cortex. Typical meristematic regions were not observed in these nodules. Based on cell sizes and patterns, a meristematic zone was thought to exist between the mantle and the inner cortex. THE DEVELOPMENTAL ANATOMY of cycad coralloid roots was most recently investigated by Milindasuta (1975). He used Korper-Kappe analysis as evidence for the presence of a persistent root cap, which was thought to form the outer covering of the entire nodule. This conclusion was also reached by McLuckie (1922) and Wittman, Bergersen and Kennedy (1965)? but was opposed to that of Schneider (1894), Life (1901), Chaudhuri and Akhtar (1931), and Schaede (1944), who did not observe a root cap in similar nodules. These latter authors thought that the algal-zone developed in the cortex. HoweverS other workers regarded the protoderm covered by a persistent root cap as the source of the algal zone (Wittman et al., 1965; Milindasuta, 1975). In all of the above studies, soil-grown roots, which frequently contained endosymbiotic cyanobacteria, were used for anatomical analysis. Since the cause(s) of nodulation was (were) uncertain, an incomplete representation ofthe developmental stages ' Received for publication 27 March 1982; revision accepted This investigation represents a portion of a dissertation submitted in partial fulfillment of the requirements for the Ph.D. degree at the University of Montana, Missoula. The author wishes to acknowledge the intellectual guidance of his research advisor in the Botany Department, Dr. David E. Bilderback. occurring during nodulation was obtained. Recently, Webb (1981, 1982a) discovered that light caused nodulation of Zamia pumila L. secondary roots in sterile culture. This led to the present study in which the developmental anatomy of Z. pumila root apices was examined during light-induced nodulation in sterile culture. MATERIALS AND METHODS-Seeds of Zamia pumila L. (-Z. floridana DC.;= Z. integrifolia Ait.) were donated by Fairchild Tropical Garden. They were prepared and cultured aseptically as previously described (Webb, 1982a). Seedling cultures were maintained in a dark incubator for 30 days at 27 C to allow for secondary root formation. Secondary roots were harvested at this time and after 5, 20, 60, or 120 days exposure to 1 2-hr photoperiods of 5 klux (General Electric Cool White fluorescent tubes). Light intensity was measured at the level of the culture tubes with a Gossens Luna Pro photographic light meter. Light and dark temperature was maintained at 27 C in plant growth chambers. For microscopic examination root tips were excised and fixed in 3% glutaraldehyde buffered to pH 6.8 with 0.05 M potassium phosphate. Some roots were also post-fixEd in 1% (W/V) OS04 buffiered to pH 6.8 with potassium

We launched imbibed seeds of Zea mays into outer space aboard the space shuttle Columbia to determine the influence of microgravity on cellular differentiation in root caps. The influence of microgravity varied with different stages of cellular differentiation. Overall, microgravity tended to 1) increase relative volumes of hyaloplasm and lipid bodies, 2) decrease the relative volumes of plastids, mitochondria, dictyosomes, and the vacuome, and 3) exert no influence on the relative volume of nuclei in cells comprising the root cap. The reduced allocation of dictyosomal volume in peripheral cells of flight-grown seedlings correlated positively with their secretion of significantly less mucilage than peripheral cells of Earth-grown seedlings. These results indicate that 1) microgravity alters the patterns of cellular differentiation and structures of all cell types comprising the root cap, and 2) the influence of microgravity on cellular differentiation in root caps of Zea mays is organelle specific.

The anatomy of Macrozamia communis L. Johnson lateral roots and nodules was studied following axenic culture in light and darkness. Pointed lateral roots from dark cultures had an open apical organization similarto that of other cycads and gymnosperms. A distinct protodermderived epidermis was not observed. At the apex, the dermis was formed by the outer root capcortical cell layer. Subapically, the outer cortex formed the dermis. No evidence of an algal zone was observed in these roots. The stele was bounded by a distinct endodermis and contained an exarch, diarch xylem. Apogeotropic nodules which developed at the root-shoot junction in darkness, branched dichotomously and had rounded tips covered by tangentially-enlarged root cap cells. The root cap was reduced to a few cell layers and was confined to the extreme nodule apex. The central region of the apical meristem was enlarged, and meristematic cells contained differentiated amyloplasts. A presumptive algal zone was present in some but not all nodules and divided the cortex into inner and outer regions. Stelar anatomy was similar to that observed in pointed, dark-grown lateral roots, except that there was greater xylem differentiation. Nodules which developed in the light were similar to dark-formed nodules, except that root cap cells were radially enlarged and extended over the flanks of the nodule forming a persistent root cap. The heteromorphic lateral roots of M. communis formed a developmental continuum not a heterorhizic root system. THE CONCEPT of heterorhizy was developed to account for extreme morphological variations associated with specialized root functions (Wilcox, 1964). The heteromorphic root systems of conifers and other woody species provided many examples which nurtured this idea. A change in the number of primary xylem poles was one feature of heterorhizy. With Libocedrus decurrens the number of xylem poles was positively correlated with increased root diameter (Wilcox, 1962a, b). Since the apical meristem is generally regarded as the site of vascular pattern determination in roots (Torrey and Wallace, 1975), changes in meristem size which accompany altered root development should be reflected in stelar anatomy. However, the basic pattern of xylem differentiation for Pinus resinosa long and short roots was diarch. In this case, Wilcox (1964) con' Received for publication 1 October 1985; revision accepted 11 February 1987. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada to DTW. 2 Current address: Forest Biotechnology Centre, BC Research, 3650 Westbrook Mall, Vancouver, British Columbia V6S 2L2, Canada. 3Current address: Biology Dept., Washington University, St. Louis, MO 63130. 4 Send correspondence to this author. cluded that the number of xylem poles was unrelated to meristem size. For this and other anatomical reasons, Wilcox (1962b, 1964) disputed the validity of heterorhizy as a morphological concept. Cycad root systems are clearly heteromorphic. Seedlings typically produce an orthogeotropic tap root which forms slender, plagiotropic laterals or tumescent, apogeotropic root nodules (Webb, 1982, 1983a, b; Webb, Nevarez, and De Jesus, 1984). In several cycads tested in vitro, light was required for nodulation (Webb, 1982; Webb et al., 1984). However, with Macrozamia communis, nodules developed at the root-shoot junction in darkness (Lamont and Ryan, 1977; Webb, 1983a). With this species, light caused increased dichotomous branching, and also enhanced the papillose appearance of nodules formed in the dark. Light also converted plagiotropic laterals into papillose, apogeotropic nodules (Webb, 1983a). Cycad root nodules branch dichotomously and are determined entities which do not revert to form typical absorptive roots. Furthermore, they are adapted to fulfill a specific physiological role in that they are the only potential sites of endosymbiotic nitrogen fixation for this taxon. Nodule ontogeny in Zamia pumila (Webb, 1983b) closely resembled Pinus strobus my

The root meristem of E. monosperma and E. chilensis possesses a central group of distinctive, large cells. These cells have large nuclei with scatter.ed heterochromatin, proplastids with no starch, small vacuoles, mitochondria, few dictyosomes and endoplasmic reticulum cisternae, and lipid deposits. Over a 24 hr labelling period, the large cells fail to incorporate 3H-thymidine, whereas cells both distal and proximal to this region do. A quiescent center which includes these large cells is present therefore. Both species have an extensive root cap, the length being contributed by mitoses in many tiers of cells distal to the quiescent center. The root cap consists of a columella and peripheral regions. Distinctive amyloplasts, an increase in the number of endoplasmic reticulum cisternae and dictyosomes, large vacuoles, and lipid deposits are characteristic of differentiated columella cells. Peripheral cells elongate, lose most of their starch, and are eventually sloughed from the root. EPHEDRA, one of three genera of the rather specialized group Gnetales, is distributed in arid regions of both the Eastern and Western Hemispheres (Foster and Gifford, 1974). A detailed summary of the structure of vegetative and reproductive organs of Ephedra has been published recently (Martens, 1971). Martens points out, however, the limited information available concerning the structure of the root of Ephedra. Although several authors (Deshpande and Bhatnager, 1961; Pillai, 1966; Lehmann-Baerts, 1967; Voronin, Voronkina and Lodkina, 1973) have described the organization of the root meristem and root cap of various Ephedra species, there are conflicting views of meristem organization. There are no ultrastructural studies of either the root meristem or root cap of this genus. This paper, which is part of a broader study of roots of non-angiosperms, (see Peterson and Brisson, 1977; Brisson et al., 1977; Peterson, Scott and Kott, 1979), considers the structure of roots of two species, E. monosperma and E. chilensis. MATERIALS AND METHODS-Seeds of Ephedra monosperma C. A. Mey. were obtained from Jakutsk, USSR while seeds of E. chilensis Miers. came from the Botanical garden, University of California, Berkeley. To ' Received for publication 20 August 1979; revision accepted 16 November 1979. This research was supported by an operating grant from the Natural Sciences and Engineering Research Council of Canada. We thank Doug Grenville for help with the plates, and Dr. R. Schmid for his help in obtaining E. chilensis seeds. obtain root tips for microscopy, scarified or intact seeds were germinated in dark under normal laboratory conditions either on filter paper moistened with distilled water or 0.5% KNO3, or in vermiculite moistened with distilled water. Only roots which had reached lengths greater than 1 cm were processed for microscopy. Light microscopy and Transmission Electron Microscopy-Ten root tips of each species were excised, fixed in 4% glutaraldehyde buffered to pH 6.8 in 0.07 M sodium phosphate buffer for 24 hr at 0 C, post-fixed in 2% aqueous osmium tetroxide for 4 hr, dehydrated in a graded acetone series, and embedded in Spurr's resin (Spurr, 1969). Thick sections were stained with 0.05% toluidine blue 0 in sodium tetraborate and thin sections with uranyl acetate and lead citrate. Histochemistry-Root caps and root meristems were examined following a number of histochemical procedures including acriflavine-HCl for polysaccharides (Culling, 1974), Sudan Black B for lipids (Eurenius and Jarskar, 1970) and aniline blue for callose (Smith and McCully, 1978). For the polysaccharide and callose tests, a Leitz epifluorescence microscope using exciter filter UG1/TK and barrier filter 400/K 436 was used. Histoautoradiography-Seedlings with extended radicles were exposed to 1.0 Ci/ml 3Hthymidine (sp act 6.7 Ci/mmol) for 24 hr and given a cold chase with 0.01 mg/ml thymidine for 6 hr. Root tips were removed from 10-15 seedlings of each species, fixed in 4% glutar

Primary roots of Phaseolus vulgaris (Fabaceae) are positively geotropic, while lateral roots are not responsive to gravity In order to elucidate the structural basis for this differential georesponse, we have performed a qualitative and quantitative analysis of the ultrastructure of columella cells of primary and lateral roots of P. vulgaris. Root systems were fixed in situ so as not to disturb the ultrastructure of the columella cells. The columellas of primary roots are more extensive than those of lateral roots. The volumes of columella cells of primary roots are approximately twice those of columella cells of lateral roots. However, columella cells of primary roots contain greater absolute volumes and numbers of all cellular components examined than do columella cells of lateral roots. Also, the relative volumes of cellular components in columella cells of primary and lateral roots are statistically indistinguishable. The endoplasmic reticulum is sparse and distributed randomly in both types of columella cells. Both types of columella cells contain numerous sedimented amyloplasts, none of which contact the cell wall or form complexes with other cellular organelles. Therefore, positive geotropism by roots must be due to a factor(s) other than the presence of sedimented amyloplasts alone. Furthermore, it is unlikely that amyloplasts and plasmodesmata form a multi-valve system that controls the movement of growth regulating substances through the root cap.

In order to quantify the ultrastructural changes associated with cellular differentiation, we have performed a morphometric analysis of the ultrastructure of the calyptrogen, columella, and peripheral cells of the root cap of Zea mays. The relative volumes of the nucleus, nucleolus, and mitochondria in the protoplasm gradually decrease as a cell moves through the root cap. The relative volume of plastids increases 240% during the differentiation of calyptrogen cells into columella cells. This increase is transient, however, since the relative volume of plastids as well as starch in plastids decreases markedly as columella cells differentiate into peripheral cells. Dictyosomes and spherosomes increase more gradually than plastids, peaking in relative volume in the innermost peripheral cells (PCI). The relative volume of the vacuome decreases as calyptrogen cells differentiate into columella cells, after which it increases during the differentiation of peripheral cells. By the time the outermost peripheral cells (PCIII) are sloughed from the cap, the relative volume of the vacuome has almost tripled. These results indicate that each cell type comprising the root cap of Zea mays is characterized by a distinctive ultrastructure. Furthermore, the ultrastructural changes associated with the differentiation of these cells are organelle specific. The results of this study are discussed relative to the function of the various cell types of the root cap. THE ROOT CAP of Zea mays is an excellent system for investigating cellular differentiation. As cells are pushed out of the calyptrogen, they sequentially differentiate into columella statocytes and, finally, into peripheral cells. Concurrently, cells assume distinctive structures and functions during the course of their differentiation and movement through the cap. Although certain cell types have been described by various authors in separate papers (Juniper and French, 1970; Juniper and Roberts, 1965; Paull and Jones, 1976; Whaley, Kephart and Mollenhauer, 1959), there has been no comprehensive study of cellular differentiation in the root cap of Zea. This paper is the second in a continuing series aimed at elucidating the structure and function of the root tip in Zea mays. In a previous manuscript (Moore and McClelen, 1983) we presented a qualitative description of cellular differentiation in the Zea root cap. The maize root cap is composed of five distinctive types of cells: 1) calyptrogen cells, which comprise the cap meristem; 2) the columella, com1 Received for publication 16 March 1982; revision accepted 17 July 1982. Thanks are extended to Jim Mauseth for his excellent technical assistance with the statistical analyses and for critically reviewing the completed manuscript. This study was supported in part by a grant from the University Research Committee of Baylor University. posed of statocytes involved in geoperception; 3) peripheral cells I (PCI), located immediately adjacent to the columella and characterized by a pronounced increase in dictyosomes and spherosomes and an absence of sedimented amyloplasts; 4) peripheral cells II (PCII), characterized by the presence of mucilage between the plasmalemma and cell wall; and 5) peripheral cells III (PCIII), highly vacuolate cells separated from the root cap (i.e., in the process of being sloughed from the cap). This qualitative study (Moore and McClelen, 1983) suggested that the spatial and ultrastructural aspects of differentiation of these five types of cells comprising the corn root cap are more pronounced than previously believed. Thus, in order to better understand the differentiation of these cells, we have quantified the ultrastructure of each of the five cell types comprising the root cap of Zea mays. MATERIALS AND METHODS-Plant material and microscopy procedures-Grains of Zea mays L. (cv. "yellow dent") were germinated in a vertically oriented cylinder of moist filter paper. Three days after germination, roots were fixed in situ in 2% glutaraldehyde in 0.05 M sodium cacodylate buffer (pH 6.9) for 2 hr at room temperature. Following three washes of 10 min each in buffer, the tissue was postfixed in buffered 2% osmium tetroxide for 2 hr. Material was then dehydrated in an ethanol series

A B S T R A C T The ultrastructure of root cap cells of cultured roots of Convolvulus arvensis L. was studied in order to relate sequential changes in the fine structure with known changes in morphology and physiology determined earlier by light microscopy and autoradiography. The rapid structural changes observed in columella cells formed from the root cap initials distal to the quiescent center can be related to the function of these cells as sensors of geotropic stimuli. Even more rapid structural changes in cells of the root cap periphery can be ascribed a functional role in protecting the columella and initial cells as the root is displaced through the soil. Mechanisms are discussed whereby cell division and differentiation in the root apex may be controlled by hormones in the apex. EXPERIMENTAL and ultrastructural research on the root cap of intact seedling roots has shown that sequential changes occur in the differentiation of geotropically sensitive cells of the root cap columella and that these cells differ physiologically from other cell populations in the root cap (Clowes and Juniper, 1964; Griffiths and Audus, 1964; Northcote and Pickett-Heaps, 1966; Dauwalder, Whaley, and Kephart, 1969; Iversen, 1969; Juniper and Barlow, 1969; Juniper and French, 1970). Previous experimental research in this laboratory has confirmed that physiologically distinct cell populations occur in the root cap of cultured roots of Convolvulus (Phillips and Torrey, 1971 a, 1972). In addition, in the development of cells of the columella from the root cap initials one of the early events appears to be the termination of any further DNA synthesis (Phillips and Torrey, 1971a). Studies on the amounts of DNA within cells of the root cap of intact roots of Zea support the idea that DNA synthesis is terminated early in cells derived from the root cap initials (Clowes, 1968; Barlow, 1970). A study of the ultrastructure of the root cap of cultured roots of Convolvulus was initiated to determine the cellular events involved in the differentiation of geotropically sensitive cells and to compare these cells with other differentiating cell populations in the root cap and in the root proper.

DNA synthetic activity in the radicle meristem of embryos of germinating seeds of the obligate root parasites, Alectra vogelii and Striga gesnerioides was followed by autoradiography of 3Hthymidine incorporation. Incorporation of 3H-thymidine occurred in the nuclei of cells destined to form the vascular tissues, ground meristem and epidermis. An analysis of the distribution of labeled nuclei demonstrated the presence of a quiescent center of 2-4 cells in the radicle at the beginning of seed germination, becoming more prominent at later stages of germination. During continued growth of the radicle which resulted in a reduction in size of the meristem, cells of the original auiescent center were activated to undergo DNA synthesis. THE ANATOMICAL CHANGES in the embryos of germinating seeds of the obligate root parasites, Alectra vogelii and Striga gesnerioides have been the subject of the first paper in this series (Okonkwo and Raghavan, 1982). This study showed that cell division activity during germination of seeds in the presence of the exudate of the host root (cowpea) was mostly confined to the radicle end of the embryo. However, due to the microscopic size of the radicle and the limited number of actively dividing cells present in this organ, the meristematic region could not be delimited very precisely during the different phases of germination. Therefore, an autoradiographic analysis of DNA synthesis in the embryos of germinating seeds of A. vogelii and S. gesnerioides is of interest to follow the organization of the root apical meristem and evolution of the quiescent center. Studies on the development of the quiescent center in the radicle meristem of early stage embryos are few, although several workers have examined the origin of the quiescent center in the embryonic radicle of germinating seeds. The consensus of current opinion is that in germinating seeds, the radicle emerges without a well-defined quiescent center, but it acquires one soon after (Clowes, 1958, 1978a; 1 Received for publication 22 August 1981; revision accepted 16 March 1982. 2 This work was done while S.N.C.O. was on sabbatical leave during the 1980-.81 academic year, and was supported by a grant from the Research Committee of the College of Biological Sciences, The Ohio State University. Permanent address: Department of Botany, University of Nioerrtia NJcil-lkk a.peric. Alfieri and Evert, 1968; Jones, 1977). Even in the radicle of mature embryos of Glycine max (Miksche and Greenwood, 1966) and Malva sylvestris (Byrne and Heimsch, 1970), where a quiescent center has been claimed to be present right from the start of germination, the evidence is considered to be equivocal (Clowes, 1978a). Based on mitotic activity and staining reactions, some evidence has been presented for the existence of a group of cells analogous to the quiescent center in the radicular tip of developing embryos of certain plants (Sterling, 1955; Rondet, 1962; Vallade, 1972). This observation was confirmed by Jones (1977) and Clowes (1978a, b) who followed the distribution of labeled nuclei and mitotic figures in the embryos of Pisum sativum and Zea mays, fed with 3H-thymidine. In both plants, a quiescent center was first noticed in embryonic radicles with well-developed root cap initials or a primitive root cap. As the embryos matured, there was a decay of meristematic activity throughout the radicle, but following germination the quiescent center reappeared. The presence of undifferentiated heart-shaped embryos in the mature seeds of A. vogelii and S. gesnerioides makes them particularly suitable to follow the development of the quiescent center in the radicle meristem of early-stage embryos and of germinating seeds. MATERIAL AND METHODs-Seeds of Alectra vogelii and Striga gesnerioides from the 1972-73 harvests used in the anatomical study were used in this investigation. Methods for inducing germination of seeds and the conditions for germination and seedling growth were

Cells in mature embryos and stationary phase (SP) root meristems of pea arrest in Gl and G2 of the cell cycle. The patterns of distribution of G2 nuclei in radicles and SP meristems, with and without G2 factor, were compared by using cytophotometric analysis of the relative amount of DNA/nucleus in sectioned material. Radicles and SP meristems-were each divided into 5 zones and the ratio of G1 to G2 nuclei was determined for each zone. The G2 population in the radicle is restricted mainly to the embryonic cortex. A small part of the G2 population was located in the central cylinder and the root cap. In SP meristems without G2 factor, the pattern of distribution of G2 cells was similar to that in radicles. SP meristems with G2 factor contained G2 arrested nuclei in all regions of the root tip. In each region the percentage of G2 nuclei was higher than that in the same region of SP meristems without G2 factor. This indicates that the population of cells that responds to G2 factor is distributed throughout the root tip. AS THE EMBRYO in a seed matures and approaches dormancy, cells in the embryo stop proliferating and accumulate in the presynthetic (GI) and the premitotic (G2) periods of the cell cycle. Microspectrophotometric studies of the relative amount

Peripheral root cap cells can be isolated by gentle agitation of the root in water for less than 60 sec. Filtration of the detached cells through a 30-Am mesh screen results in a suspension of single cells. Yields of peripheral root cap cells from 28 species in 10 families ranged from fewer than 10 per root from tobacco to more than 5,000 per root from pine. The viability of freshly isolated cells from most species, including legumes and cereals, was 85% to 95%. Cultured cells from pea and soybean exhibited cytoplasmic streaming for more than a month. When incubated at an initial density of 105 cells per ml of liquid B5(P) medium, 6% to 8% of isolated pea root cap cells divided and grew into callus.

Roots of Marsilea vestita ranging from 1-120 mm in length, as well as root primordia, were analyzed to determine mitotic activity and ploidy levels in the apical cell, five well-defined regions of the root proper, and two regions in the root cap. The mitotic index of the apical cell tended to be above the overall mean mitotic index for the entire apical meristem. No diurnal rhythm in mitotic index was apparent. The cell-cycle duration of the apical cell ranged from 12.1-25.2 hr, that of other regions of the root from 16.1-41.5 hr. There was no indication of polyploidy in any part of the apical meristem except in a few procambial cells. Thus, the results support the classical concept that the apical cell is the ultimate source of cells in the root. THE CLASSICAL CONCEPT of the role of the apical cell found in many bryophytes and lower vascular plants, namely that whenever a single apical cell is present, it divides regularly and is the source of all other cells in the shoot or root, was deduced from careful analysis of the arrangement of the cells surrounding the apical cell (Naigeli, 1845; Hofmeister, cited in Nageli and Leitgeb, 1868). This concept seems to have remained unquestioned for many years although, according to D'Amato (1975), "in 1867 already, Reess had stated that growing shoot apices of Equisetum may show a discontinuous meristematic activity of their apical [cell]" (D'Amato, 1975, p. 321). However, I have been unable to find any statement to such effect in Reess' paper. On the contrary, he wrote, "Bei Allen [Equisetum Arten] wird das Scheitelwachstum vermittelt durch ununterbrochen wiederholte Theilungen der dreischneidigen, einem tetraedischen Kugelausschnitt ungefahr entsprechenden Scheitelzelle mittelst gerader, den Seitenwanden der Scheitelzelle paralleler Hauptwande" (p. 14-"In all species of Equisetum, apical growth is caused by continuously repeated divisions of the apical cell, which has three cutting faces and approximately the shape of a tetrahedral segment of a sphere; the divisions result in the formation of straight walls parallel to the lateral faces of the apical cell."). ' Received for publication 5 June 1980; revision accepted 29 August 1980. This paper represents a portion of a dissertation submitted to the University of California at Davis in partial fulfillment of the requirements for the Ph.D. degree. I am deeply indebted to Dr. E. M. Gifford, Jr., for suggesting this topic for my research and for his invaluable advice and guidance. I am also grateful to Drs. T. L. Rost and R. H. Falk for their help with various aspects of this study and to Sigma Xi for a Grant-in-Aid of Research which supported part of this work. In 1953 Buvat and Liard reported that the shoot tip of Equisetum arvense has a mitotically inactive apical cell and that it presents a typical example of a me'risteme d'attente subtended by an anneau initial. In a subsequent paper, Buvat and Roger-Liard (1954) described mitotic activity in the apical cell of the roots of E. arvense somewhat ambiguously, stating that the apical cell "est loin de proliferer en manniere continue" (p. 1258"is far from proliferating in a continuous manner") but also, "La 'cellule apicale' et ses voisines immediates se divisent sensiblement au meme rythme; il n'y a pas de territoire inactif comme dans les racines d'Allium ou de Triticum" (p. 1257-"The apical cell and its immediate neighbors divide in the same rhythm; there is no inactive region as in the roots of Allium or Triticum."). D'Amato and Avanzi (1965, p. 392), in presenting evidence for quiescence of the apical cell, quoted only the former statement. It may be noted here that the classical theory does not demand continuous activity of the apical cell. Indeed, Nageli and Leitgeb (1868) observed that in some cases, the cells in the first, second, third, and even fourth packets of derivatives (merophytes, using the term adopted by Gunning, Hughes, and Hardham, 1978) proximal to the apical cell divide even more actively than the apical cell itself. D'Amato and Avanzi's much cited contention that the apical cell is quiescent was based mainly on determinations of the DNA content of the apical cell as compared to that of nearby cells in mitosis and on labelling indices in plants fed with 3H-thymidine for periods up to 3 days in roots of Marsilea strigosa (D'Amato and Avanzi, 1965; Avanzi and D'Amato, 1970), Ceratopteris thalictroides, Polypodium aureum, and two species of Blechnum (Avanzi and D'Amato, 1967) and shoots of Equisetum

THE GROWING ROOT TIP has long been considered a favorable object for experiments on growth and differentiation. In contrast to the shoot it has a relatively simple structure, unencumbered with foliar and floral appendages. The literature on this subject has been well reviewed in the recent papers of Erickson and Sax (1956a, 1956b). In 1939 Sinnott pointed out the suitability for growth studies of the roots produced by certain species of small-seeded grasses. The transparency of the fine roots and the rudimentary character of the root cap permit unobscured microscopic observation of the surface of the living meristem. Investigations on the growth of one of these species, Phleum pratense L., were reported by Sinnott and Bloch (1939a) and by Brumfield (1942).

IN HIS OWN researches (Thomson and Sifton, 1925; Sifton, 1940) and in a recent review Sifton (1944) has pointed out the merits of the experimental approach to studies in developmental morphology. One matter of first importance in this mode of attack is the suitabilitv of the experimental material. In this regard, few cells lend themselves so well to an intensive developmental study as those of the root epidermis. Their superficial position and arrangement in longitudinal rows enable the investigator to study them from the time of their inception in the meristem to their final condition. An additional interest is provided by their capacity to push out sideways to form hairs. During the course of an experimental study on root-hair development, the present writer (1935, 1937, 1944, 1945) has worked with two different kinds of roots, namely, those in which the epidermis is differentiated into two different kinds of cells and those in which the cells are all of one kind. For example, in cabbage and white mustard roots (Cormack, 1935) the epidermis is differentiated in the meristem into alternating rows of short and long cells. The criterion used in distinguishing the two cells is their length as seen in surface view. Also the short cells appear to vacuolate much later than the long cells and to a lesser extent. Under most conditions the short cells at maturity form hairs while the long cells remain hairless. Hairs are densely developed in aqueous solutions within a comparatively wide range of pH. On the other hand, the epidermis of tomato roots is not differentiated into short and long cells. Although they are all long they vary in length in the same region under the same conditions. Because of their length and highly vacuolated condition at an early stage, it was suggested in the last paper (Cormack, 1945) that they correspond to the long cells of cabbage and white mustard roots. This viewpoint is substantiated by the relatively poor development of hairs in aqueous solutions and extreme sensitivity to changes in pH. The present paper reports the results of a developmental study of epidermal cells in both white mustard and tomato roots. It is hoped that the work reported here will help to uncover the cause for differences between short and long cells in white mustard and for the undifferentiated condition of the epidermis in tomato roots. MATERIALS AND METIHODS.-Seeds of tomato (Lycopersicum esculentum Mill) and white mustard (Brassica alba L. Boiss.) germinate readily between moist blotters at room temperature (220C.). 1 Received for publication January 15, 1947. I.wish to thank Professor R. B. Miller of the Department of Zoology for taking the photomicrographs. When the roots reached a length of about 1 cm., they were killed in CRAF and embedded in paraffin. In order to study the epidermal cells in relation to neighboring cells the roots were sectioned transversely at 10 / beginning at the extreme apex of the root. Longitudinal sections were made also of several roots in testing various points of interest observed in transverse section. Later, corn roots (Zea Mays L.) were used in checking up the hypotheses that grew out of the investigation. RESULTS.-As in most roots, the extreme apex of both mustard and tomato roots, exclusive of the root cap, presents in transverse section a circular compact cushion of meristematic tissue, which as it increases in width at successively higher levels gives rise to stele, cortex and epidermis. Photomicrographs (fig. 1-6) illustrate the main features exhibited in transverse section at certain levels of development in the growing point of each kind of root. Although the numerical values recorded at this time are applicable to the root in question, they are typical of the results consistently obtained. Fig. 1-3 represent transverse sections made at successively higher levels in the growing point of a single white mustard root. Fig. 1, through the actively meristematic region of the root some 410 pt from the root apex, shows the section surrounded by two rows of small transparent root cap cells. Within the root cap lies the single row of epidermal cells. This row of cells, which is first distinguishable at about 130 ,u from the root apex, now shows at this higher level (fig. 1) considerable differentiation. In this row the wedge-shaped cells full of protoplasm and with a prominent nucleus are the short cells, while the radially elongated cells situated between two short cells are the long cells. As is clearly evident the long cells have begun to vacuolate. The position of the short cells, each in contact with two cortical cells, also serves to distinguish them. Between the developing epidermis and stele lie the cells of the cortex, easily distinguishable because of their large size and uniform arrangement in opposite rows. Differences in cell arrangement have, of course, been noted by many investigators. Sinnott and Bloch (1941) emphasize this feature in their consideration of the relative position of cell walls in developing plant tissues. In this study they recognize two main types of primary tissue arrangement, namely, opposite and alternate. In the present study the former is well exemplified in fig. 1-3, and the latter in fig. 4-6. Another feature of considerable interest at this stage (fig. 1) is the occurrence of intercellular spaces. According to Priestley (1929) as a dividing cell begins to vacuolate the cell wall becomes more elastic and offers more resistance to the outward

THE SYNTHESIS of starch, the most important reserve carbohydrate in plants, is a problem which has commanded considerable attention. A major advance in understanding starch transformations was the discovery by Hanes (1940a, b) that the linear fraction of starch may be synthesized in vitro by phosphorylase prepared from peas or potatoes. Later, an enzyme concerned with the synthesis of the amylopectin component of starch was reported by Haworth et al. (1944), Bourne and Peat (1945), and Bernfeld and Meutemedian (1948a, b). The mechanism of action and properties of purified phosphorylases have been discussed in a number of excellent reviews (Cori and Cori, 1946; Hassid, 1946; Haworth, 1946; Kalckar, 1945; Peat, 1946; Wynne, 1946; etc.) Certainly more is known about the in vitro reaction of starch synthesis by cell extracts than its synthesis within the extremely complex environment of the plant cell; the reactions occurring in vivo may not be entirely comparable to reactions with cell extracts. Yin and Sun (1947) developed a histochemical stainino procedure by which the presence of phosphorylase can be detected in plant tissue. Using this technique, they studied its physiological role in the living plant by correlating the distribution of the enzyme with the site of normal starch synthesis in germinating soy beans (Yin and Sun, 1949). Still more should be done to elucidate the role of phosphorylase in plant cells, and it was the purpose of the present work to investigate further some aspects of starch synithesis in vivo and in situ. MATERIALS AND METHODS.-From a variety of plant parts examined for phosphorylase activity, the primary root tips of pea seedlings were selected as the test material. They demonstrated fairly consistent activity, and the presence of stored starch, which occurred only in the root tip of very young seedlings and to a limited extent in the root cap, was not a problem. The pea seeds (Wisconsin Early Harvest, Olds Seed Co.) were soaked and aerated, germinated between damp paper towels until the radicles were 2-3 cm. long, and then transferred to aerated mineral solutions in the dark. Seedlings 10-14 days old were usually used. The method of Yin and Sun (1947) in which free-hand sections are incubated in buffered glucose-l-phosphate solution was modified for many of the experiments bv cutting longitudinal sections 40-60 V. thick on a freezino microtome. When it was desirable to avoid injury to the cells from freezing

KNOWLEDGE CONCERNING the processes of growth and division of individuLal meristematic cells as they develop to maturity is clearly of importance for an understanding of the factors involved in cell differentiation and plant morphogenesis. Little is known about these processes in multicellular tissues except from fixed and sectioned material, the limitations of which are obvious. The most significant evidence can be obtained only by the use of living tissue wherein the same cell can be observed throughout its development. Recently a technique has been devised whereby growth and division of single cells in living root meristems can -be studied. Results obtained by the use of this method are reported in this paper. Sinnott (1939) has pointed out that the epidermal cells in root meristems of many small-seeded grasses can be clearly seen in the living condition since the root cap consists of but a few loose cells at the extreme tip. Seedlings are grown on microscopic slides covered with moist lens paper and kept in covered Coplin staining jars with a little water at the bottom of the jars. The growing seedling is flooded with water, covered with a supported cover glass, and placed under the microscope for observation of the growing root. By means of camera lucida drawings made at successive intervals of time, it is possible to observe cell division and growth. Using this technique, Sinnott and Bloch (1939a) made a study of root hair formation in Phleum pratense, Poa trivialis, Sporobolus cryptandrus, and Chloris gayana. It was found that in Phleum and Poa the last cell division is an unequal one which results in a small trichoblast or root hair-forming cell and a larger cell which does not form a root hair. The ,trichoblast is always toward the tip of the root and the hairless cell toward the base. During elongation, the trichoblasts grow less than the hairless cells. In Sporobolus and Chloris the divisions were found to be more nearly equal and both daughter cells may form root hairs. In a later paper (Sinnott and Bloch, 1939b), especial attention was given to "sliding growth" and differential wall growth, using Phleum and Sporobolus for observation. All the evidence obtained indicated that sliding growth does not occur. Differential wall growth was apparent in that: (1) parts of the walls of hairless cells opposite trichoblasts grow less than parts opposite other hairless cells; (2) under certain conditions, cells grow more in the middle portions than near the ends; (3) the end of a cell toward the base of the root attains its final size sooner than the apical end of the same cell. 1 Received for publication March 23, 1942. The author wishes to express his appreciation to Professor E. W. Sinnott of Yale University for criticisms and suggestions during the course of this investigation. This paper is part of a dissertation presented to the Graduate School of Yale University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. These observations show that the differences in size of large cells and small ones are a result not only of unequal divisions but also of differences in their growth rates. The present investigation was undertaken to determine the growth rates of single cells and their descendants as they pass from the meristematic condition to maturity, thus obtaining more

DAUGHERTY, LYMAN H. (U. Michigan, Ann Arbor.) Triassic roots from the Petrified Forest National Park. Amer. Jour. Bot. 50 (8): 802-805. Illus. 1963.-Araucariorhiza joae, a new genus and species of Triassic roots, is described. The description includes the root tip with a columella type embryonic region, the secondary roots, the small tertiary roots, and a discussion in regard to the relationship of this form genus. THE FOLLOWING description is based on fossil roots collected by Chief Naturalist Philip F. Van Cleave and on subsequent collections made by Mr. Van Cleave and the writer. The material had eroded from a sandy bentonite of the Chinle formation and a number of specimens were found in a small area at the Itopsidema locality (Daugherty, 1960). This locality is situated approximately 4 mile due north of Agate Bridge in the east central part of Sec. 4, T. 14 N., R. 24 E., within the boundary of the Petrified Forest National Park.2 GENER?AL FEATURES-The fossil roots were discovered within large fragments of a hollow log of Araucarioxylon arizonicum. The log was approximately 35 cm in diameter and the entire hollow center was filled with small roots. The secondary roots are ca. 1-4 mm in diameter and the tertiary roots are ca. 0.4 mm in diameter. Apparently the log had been in a spongy, water-soaked condition before petrifaction occurred, for small branch roots from the central portion of the log had also -penetrated the wood (Fig. 3). Thin sections show perfect preservation in many of the roots, and, judging from the color, the cell walls had retained a considerable amount of carbon after the impregnation of silica formed chalcedony. THE ROOT TIP-A longitudinal section of the tip of a secondary root 630,u in diameter, and with excellent cell structure, proved to have a meristematic region common to many extant gymnosperms (Fig. 1). In many, if not all, of the recent conifers the root cap contains a central columella consisting of transversely dividing cells that also can form daughter cells laterally as described by Esau (1960) for Picea and by De Bary (1877) for Juniperus. The fossil root tip contains a distinct columella, and it is assumed cells of the root were formed in the same manner as those being formed by roots with a columella at the present time. I Received for publication January 26, 1963. The writer is indebted to the Petrified Forest National Park Corporation for expenses incurred on trips to the Petrified Forest, and to the National Science Foundation for financial assistance under grant G-19553, C. A. Arnold project director. 2 The Petrified Forest National Monument became a National Park on December 8, 1962. TERTIARY ROOTS-Small branch roots that were apparently modified for the absorption of water and salts in solution were featured by bulb-like expansions near their tips (Fig. 2, 4). No root caps or root hairs were observed, and epidermal cells of the bulb-like expansions possibly served to carry on absorption. It is not unusual for epidermal cells to function in this manner, for in recent roots having root hairs water absorption may be accomplished by ordinary epidermal cells and is not restricted to the root hairs (Esau, 1960). The stele of the roots is surrounded by an endodermis having Casparian bands and the cortex consists of parenchyma. The stele is diarch, and the cambium was inactive or absent, and, like many small branch roots of extant species, no secondary xylem or phloem was formed. SECONDARY ROOTS-The primary xylem of the secondary roots is essentially like that of the tertiary roots. However, all secondary roots developed root caps and a row of cambial initials on both sides of the metaxylem. When the 2 rows of cambial initials were active, 2 cuneate groups of secondary xylem were formed that appear faiishaped in transverse sections (Fig. 7, 8). Often only 1 row of cambial initials became active, and 1 "fan-shaped" structure of secondary xylem was formed (Fig. 6). When a single "fan" developed, the primary xylem was displaced and forced from the center of the root toward the periphery. The secondary xylem is composed of tracheids with 2 rows of alternating bordered pits on their radial walls (Fig. 5), and, in addition to the typical araucarian pitting, a number of tracheids apparently contain resin plugs. Xylem rays, 2-6 cells high, are uniseriate and consist of procumbent parenchyma cells. In many roots, the phloem is not well preserved, but occasionally it shows excellent preservation (Fig. 6). Prosenchymatous cells ca. 600,u long and 25,u wide are evidently sieve cells although no sieve plates can be observed. Sieve cells of the phloem are approximately square or slightly rectangular in the transverse sections. They are associated with phloem parenchyma and a few isolated sclerotic cells were observed in the periphery of the phloem.

In cap cells of intact plant roots exposed to lmM uranyl for 30 min or less, uranyl crystals were found only in cell walls and in secretory products which had been extruded from the protoplast. In roots exposed for 10-20 hr to 0.1mM uranyl, packets of uranyl crystals bound to secretory products were found within the protoplasts of those exterior cells which contained accumulations of secretory products between the cell wall and protoplast. Although the evidence indicated that these packets of crystals entered the protoplast pinocytotically, results with these specialized exterior cells did not apply to the vast majority of root cap cells in which, after prolonged exposure to 0.1mM uranyl, crystals were concentrated in vacuoles. In roots exposed to 1 or 5mM uranyl for 1 hr, the plasmalemma of interior cap cells was much thicker (13.1 nm) than normal (8.2 nm), and many invaginations and vesicular structures were found near the protoplast surface. Crystals were confined to cell walls except for a few found in vesicles with thickened membranes. Serial sections indicated that most vesicular structures with thickened membranes were in contact with the cell wall, but a few, including some which contained uranyl crystals, were within the protoplast. These results provide evidence of pinocytotic activity in intact plant cells exposed to a toxic heavy metal. PINOCYTOSIS, in the general sense of a process by

The duration of the subdivisions of the nuclear cycle in cells from the root tips of Tradescantia paludosa was determined by the labelled mitosis method. These times were found not to increase or decrease in direct proportion to each other at the three temperatures tested, 13, 21 and 30 C. The main difference between the nuclear cycles at 30 and 21 C was in the shortening of the pre-DNA synthetic period (G1) at the higher temperature. A comparison between the nuclear cycle at 21 and 13 C showed that at the lower temperature mitosis and the post-DNA synthetic period (G2) were approximately tripled in duration while DNA synthesis was doubled. Cell synchronization and radiosensitivity are discussed in relation to these findings. THE EFFECT of the environment on the girowth of cell populations has long been of interest and numerous studies have been performed in attempts to determine changes that may occur in populations grown at different environmental means. A factor that changes almost continuously under natural conditions is the temperature, and it is without doubt one of the major items that influences the rates at which cell populations grow (Brown, 1951; Evans and Savage, 1959; Hughes, 1952). This paper is devoted to a description of the growth of cells in root tips of Tradescantia paludosa at three different temperatures. The popuilation is analyzed by using the labelled-mitosis technique (Wimber, 1960) with tritiated thymidine and is subdivided with respect to the duration of the DNA synthetic period and mitosis. It is shown that the durations of the various subdivisions of the nuclear cycle are bDy no means uniformly lengthened or shortened by temperature changes, but some parts are altered much more profoundlv than others. MIATERIALS AND MIETHOI)S-Root tips of Tradescantia, paludosa were used as experimental material. Plantlets were removed from the clonal parent (B2-2, A. H. Sparrow's clone) and the bases placed in aerated Hoagland's solution. They were given constant light (ca. 500 ft-c) and constant temperature. Three temperatures were used, 13, 21 and 30 C. Well-developed roots were usually present within a week. Nuclei in the root meristems were labelled by growing the rooted plants in an aerated nutrient solution containing 0.5 ,c/ml H3-thymidine (New England Nuclear, 3.6 c/mmole) for 30 min .The roots were washed thoroughly in water following label' Received for publication February 15, 1965. Part of this study was carried out at Brookhaven National Laboratory under the auspices of the U. S. Atomic Energy Commission, and part was supported by Public Health Service Research Grant GM 11702-01 from the National Institute of General Medica.l Sciences. The technical assistance of Miss Mary Winnike is gratefully acknowledged. ling and placed in a nutrient solution for further growth. Root tips were collected at various intervals after labelling and fixed in acetic-alcohol (1:3); they were hydrolyzed for 10 min in N HCI at 60 C and stained by the Feulgen method. The root cap was removed and squashes of the intensely stained terminal portion of the root tip were prepared on slides; the cover glasses were removed by the method of Conger and Fairchild (1953). Autoradiographs were made using Kodak NTB liquid emulsion. Exposure was for approximately 3 months. The percentage of mitotic cells labelled was determined by using samples of 100 prophase cells from each slide. Eight roots were sampled at each time interval. Ratios of labelled cells to cells in mitosis were also determined in some cases, and with these a total of 1000 cells were scored on each slide. The labelled-mitosis technique is the method utilized here (Quastler and Sherman 1959; Wimber, 1960, 1963; Evans and Scott, 1964) for the determination of the nuclear cycle and its various subdivisions as they relate to DNA synthesis. The nuclear cycle (C) for this type of analysis may be divided into mitosis (M), DNA synthesis (S), pre-DNA synthesis (G1) and post-DNA synthesis (G2). A discussion of this method is given in the above references and need not be repeated here. Suffice it to say that the duration of the subdivisions of the cycle may be estimated from a labelled prophase (P) curve. An example is given in Fig. 1, and estimates of the durations of G2 + P/2, S and C may be read directly from the figure as distances on the abscissa, using as end points the positions at which the rising and descending curves cross the 50% mark. The duration of GI + mitosis P/2 may be calculated by subtraction of G2 + P/2 + S from C. The duration of mitotic stages may be estimated by observing the frequency of labelled cells and mitotic figures in root tips fixed immediately after a 30-min labelling period. In an asynchronous

A B S T R A C T The durations of the cell cycle in physiologically different regions of the meristem of cultured roots of Convolvulus arvensis were determined by the metaphase-accumulation technique involving colchicine. The cell cycle in the root cap increases from 13 hr in the actively dividing initials of the first tier to 155 hr in the slowly dividing initials of tiers 2-4 to an indeterminate value for derivatives of the initials in the root cap columella. The cycle times for the cells of the central cylinder and cortex are 21 and 27 hr, respectively. The cells of the quiescent center have a cycle of an estimated 420 hr. The duration of the cell cycle in these different regions is discussed in relation to the increased duration of G1 in slowly or non-dividing cells. The possible regulation of cell division by the synthesis of a cell-division factor in the quiescent center is also discussed. EARLY STUDIES on the duration of the cell cycle in root meristems of intact plants yielded average values for the total elapsed time from one mitosis to the next for cells of the whole meristem. These studies did not consider possible differences in cycle times of populations of cells within the apex which might be physiologically distinct from one another. Then Clowes (1956) showed that a region of infrequently dividing cells, which he called the quiescent center, existed in close proximity to rapidly dividing initials which gave rise to different populations of cells in the apical meristem of roots of Triticum and Zea. Thereafter, attention shifted to the determination of cell-cycle times in different regions of the apical meristem in an attempt to understand how differential rates of cell division influenced the development and maintenance of an organized meristem. Clowes (1961) determined the total length of the cycle for cells of the quiescent center, the root cap initials, and the initials of the stele and cortex in seedling roots of Zea mays by two different experimental techniques; namely, continuous labelling with tritiated thymidine, and metaphase accumulation with colchicine. He reported that the root cap initials had the shortest cycle (12 hr), while cells of the quiescent center had the longest (239 hr). The initials of the stele and cortex had cycle times of 25-28 hr and 48-55 hr, respectively. Using the pulse-labelling technique with tritiated thymidine, Clowes (1965) found that the duration of the cycle for the root cap initials and the stelar initials in roots of Zea mays was 14 and 22 hr, respectively. Also, using the pulse

A B S I R A C T Cells in the interior of susceptible oat roots treated with the disease-inducing agent victorin exhibit many of the ultrastructural features which characterize the epidermal or outermost root cap cells of untreated roots. An increase in electron density of cell walls fixed in permanganate is the first effect of victorin seen in the root interior. Other early victorin-induced changes are formation of enlarged, densely stained vesicles by the Golgi apparatus and organization of the endoplasmic reticulum into roughly parallel profiles. All of these features are characteristic of untreated epidermal cells. V'ictorin also induces the formation of large numbers of lomasome-like wall lesions and causes a marked increase in the number of nearly spherical, membrane-bounded structures tentatively identified as spherosomes. Similar lomasome- and spherosome-like structures are much more abundant in the outermost cells of the root cap than in other regions of untreated roots. This suggests that these structures may be characteristic of cells destined to undergo disintegration. Victorini-induced lesions appear to arise within the cell wall as the result of an activation of wall-degrading enzymes. An early change which makes the unit structure of the plasma membrane visible over extended areas may account, for victorin-induced changes in permeability. Disrupted plasma membranes and swollen mitochondria are found only in cells heav ly damaged biy victorin. Mfany of the effects of victorin resemble those of calcium deficiency and calcium is known to suppress victorin-induced disease symptoms. This suggests that calcium nutrition may play a role in the pathological changes induced by victorin.

Analysis of incorporation of labeled precursors into DNA coupled with mitotic counts furnishes a picture of the sequence in which tissues become activated during seed germination. By appropriate sampling techniques ontogenetic changes within an organ during germination can be traced and correlated with anatomical information already available. This information will aid in making physiological studies during this phase of plant development more meaningful. In this preliminary study it was found those tissues which mature during embryogeny (root cap, coleorhiza, scutellar node) are first to resume DNA synthesis (30 hr after soaking). No mitoses were observed in these tissues. In shoot and root, mitoses usually precede DNA synthesis, indicating that the nuclei of the dormant embryo have a DNA value of 4C or more. The shoot begins its activity much later than the root (50 hr). The shoot apex was the last region to become active, some 70 hr after initiation of the soaking treatment. (auth)

root cap, and characterized by relatively slow elongation, absence of cell division and, hence, the presence of vacuolated, extending cells. Cell differentiation occurs at different levels within the root, depending upon the particular type of cell involved. Thus, sieve tubes are matured within 230 ,u of the base of the root cap, well within the meristematic zone; while the youngest xylary elements to show characteristic thickenings of the secondary walls are found at a point 740 p further back, in the zone of cell elongation. For this reason, the use of the terms "zone of differentiation" and "zone of maturation," without designating a specific cell type, should be discontinued.

books usually figure the root tip in longitudinal section with various regions marked off: root cap, meristematic zone, zone of elongation, and zone of differentiation or maturation. Such descriptive demarcations fail to give quantitative information concerning the absolute or even the relative rates at which the various developmental processes, which are supposed to be occurring in these regions, are actually taking place. The nature of the transitions between one zone and the next are often completely ignored. 1Received for publication July 14, 1914. Growth rate measurements were made by the junior author, who is now serving with the armed forces. In 1939, Sinnott pointed out the suitability of certain species of small-seeded grasses for, studies of the growing root tip. In these species the transparency of the fine roots and the rudimentary character of the root caps permit unobscured microscopic observation of the surface of the living meristems. Brumfield (1942) has made some interesting experimental studies on one of these species, Phleum pratense L., in which rates of elongation of roots at various distances from the root apex were measured. In this paper methods are developed for determining rate of elongation of a root, cell elongation, rate of transverse cell division, and rate of increase in cell wall area, each expressed as a function of distance from the root apex. These methods are then applied to roots of Phleum pratense. The resulting data give a graphic picture of the above-mentioned developmental processes which are taking place within the growing region of a root and of the complex interrelations between them. Some further observations on the differentiation of vascular elements are correlated with these findings. THEORETICAL METHODS.-Rate of elongation.Seedlings of grasses such as Phleum pratense may be grown in vertical culture chambers so constructed as to allow examination of the root meristems under high magnifications with a compound microscope. It is possible to measure the amount of growth occurring during short time intervals between the root apex and characteristically-shaped epidermal marker cells. When such growth increments in unit time (r) are plotted against the initial distance (x) of the marker cells from the root apex, sigmoid curves, representing the rates of displacement of x from the root apex, are obtained (fig. 2). It should be pointed out that the time interval must be sufficiently short to prevent undue displacement of the marker cell from its original position, x units of distance from the root apex. The first derivatives of these sigmoid curves, dr/dx (the slope of the tangent to the curve at any point), give the rates of elongation of the roots as a function of distance from the tip (fig. 3). Average length of epidermal cells.-The length (1) of epidermal cells at various distances from the root apex may be readily determined. Camera lucida drawings of the surface of actively-growing roots cannot be made with sufficient rapidity to serve for this purpose. If large numbers of cells are to be measured, it may be convenient to make such measurements upon roots killed in fixatives which do not appreciably distort the cell dimensions, although measurements from photographs of living roots should be preferable. The length of cells in the subepidermal layers may be studied from longitudinal sections. Figure 4 shows the average length of epidermal cells with their basal ends at x.

Stem branches in Quercifilix arise in the axils of roots at their attachment to the parent rhizome or apically upon a new root before it has erupted from the tissues of the rhizome. In either case, they arise from the surface cells of the accessory calyptra. When the new stem arises apically, the root apex and the root cap give rise to a continuous core of procambium. The apical cell of the root becomes a tracheid, more or less retaining its original shape. The rootshoot axis thus produced is referred to as a rhizoclad. THE RHIZOCLAD, a single continuous axis with root at one end and stem at the other, is a botanical curiosity in taxa that show a sharp differentiation between roots and stems. The one described below was discovered quite by accident during a study of fern stem apices. MATERIALS AND METHODS-The plant material was obtained from cultivation. Its preparation involved the use of standard histological procedures. Staining was done with Heidenhain's hematoxylin and fast green. DESCRIPTIONLateral buds in Quercifilix (Fig. 1 ) on the rhizome replace roots in a regular system of rhizotaxy. They arise from an irregular surface meristem on the accessory calyptra of certain roots. The accessory calyptra is that covering of an endogenous root tip that arises from the epidermis and outer cortical layers of the parental organ, in this case rhizome. Ferns in general produce accessory calyptras that are ruptured eventually by greater growth of the root apex below. The amount of accessory calyptral growth determines the extent of the collar of torn tissue present at the base of the older root. It might be guessed that cells external to an endogenously produced root apex divide anticlinally as a response to physical pressures from below, an interpretation perhaps valid in certain instances, but at the surface in Quercifilix the surface meristem of the accessory calyptra is composed of a number of small cells and is often somewhat sunken and overgrown from the sides. ] Received for publication 29 January 1985; revision accepted 30 April 1985. 2 Present address: Rt. 3, Box 188, Picayune, MS 39466. Certain root tips are broader than others, even before they elevate the surface of the rhizome. The broader root tips grow, and their accessory calyptras grow with them. When they protrude at the surface they expand laterally, become often near-spherical and quite green (Fig. 5). Much of the lateral tissue of these expanded bodies is derived from the accessory calyptra, and, in fact, such root tips are meristematically quite inactive with activity restricted to the accessory calyptra. The specimen in Fig. 5 is actually a nearly completely formed rhizoclad. It shows a stem apex overlapped by scales at the top and a procambial connection between the root stele and the stem stele. The root stele curves sharply out of the plane of section. A younger rhizoclad is shown in Fig. 3, where only the central core of tissue appears. A welldifferentiated stem apex is at the distal end of the rhizoclad (right) and some cells (ca. 10-12) below is the apical cell of the root. Differentiation of procambium in the root had already progressed so far that the root apical cell was quite procambial in nature. In older specimens in which the stelar tissues are mature, the apical cell can usually still be recognized as a tracheid. The specimen in Fig. 2 is more advanced than those in Fig. 3 and 5. The stelar tissues of this rhizoclad are well laid out as procambium. The narrow root stele appears within the tissues of the parental rhizome (only cortex of which is shown) and the expanded stele of the new rhizome on the right in the distal end of the rhizoclad. The scales covering the new stem apex were removed in order to identify the structure before fixation. The rhizoclad in Fig. 1 is several times longer than the one of Fig. 2. It is approximately one

One can find in various texts the statement that the root-cap cells of plants die and are sloughed off, and it is probably the general opinion among botanists that the root-cap cells are either dead when they are sloughed off or that they die soon thereafter. Thus, in Jost's Plant Physiology (translation by Gibson, p. 283) the statements concerning the rootcap read as follows: " Its cells are short-lived, but they are constantly being renewed. In spite of this renewal the root-cap does not increase in size because the older cells die off in front and are cast off as new ones are formed." That the root-cap cells, when sloughed off, are not necessarily dead or short-lived but may persist for many days, seems to be substantiated by various observations made by the writer with a number of different plants. In view of the increasing attention being devoted to the subject of root excretions, it seems desirable to make record of these incidental observations. The observations here recorded were made primarily on the root-cap cells of corn, although similar observations were made with Canada field pea. The plants were grown in water cultures under sterile conditions, that is, with the roots growing in the entire absence of microorganisms. Pfeffer's nutrient solution was used, with the replacement of dibasic potassium phosphate for monobasic potassium phosphate and with or without one half percent sucrose. The root-cap cells generally collected at the bottom of the culture vessel, appearing as slimy masses, the amount increasing with the age of the culture and the cells always being more abundant in the sucrose cultures. The cells were sometimes isolated, sometixnes in chains of from two to seven cells, again in plates, and occasionally entire root-caps were nloted. These cells were always in a healthful condition. They were well filled with protoplasm and each possessed a conspicuous nucleus centrally placed. That the sloughed-off cells are not short-lived is borne out by the following: Various corn cultures, both with and without sugar, were examined. When the cultures were forty-five days old, the precipitates in the culture vessels were examined and in no case could a dead cell be found. Every cell was well supplied with protoplasm, and that the cells were living was apparent not only from their general appearance but also from the fact that the cells could be plasmolyzed by a glycerine solution and recovery followed.