Search

1. A technique has been described for the preparation of clots from purified fibrinogen and thrombin of bovine origin which are suitable for study with the electron microscope. Experiments have been carried out to compare the fine structure of clots prepared at various values of pH. 2. The clots are composed of meshworks of single and compound fibers. At pH 8.5 the unit fibers have a smaller average diameter than those formed at pH 7.6 or pH 6.3. The tendency for the lateral association of unit fibers into compound fibers is markedly increased as the pH is decreased. 3. A striking feature of all the clots studied is cross-striation of the unit fibers. The periodicity of these striae is quite constant throughout (approximately 250 Å). There is a precise coincidence of the striations of the individual unit fibers where these are associated side by side to form compound fibers.

We have shown that a small population of normal cells can be cultured from the scales of the squirrel fish, Holocentrus rufus. They can be grown directly on Formvar-carbon-coated gold grids and, while still on the grids they can be fixed, stained and dehydrated for high voltage electron microscopy. One of the cell types (epidermal) spreads out on the carbon-coated surface and is thin enough in most parts for conventional (100kV) electron microscopy. The aspect of wholeness represented in these qells should not be overlooked for it provides information that might be missed in a series of thin sections where the sample is obviously smaller. Furthermore, if experimental studies are contemplated, they can be made while the cells are still alive and available for light microscopy.

The recent history of cell biology cannot be recorded without generous reference to electron microscopy. After all, it was this instrument and the enormous increase in resolution it provided that gave cell biologists access to information that even the most imaginative of investigators had not submitted to the printed page. This paper will record highlights that a few investigators introduced into specimen preparation. This pursuit has contributed more to the progress, i.e. history, of the field than is generally recognized. The first images of cells that can be said to reveal something new were published in 1945. They were micrographs of thinly spread cells grown in tissue culture on surfaces coated with formvar and thence with evaporated carbon. After the suitability of a cell or group of cells had been established by light microscopy, they were freed of culture medium with balanced salt solution and fixed with OsO4. Thereafter, and before drying, the selected cells were washed in distilled water, and to get the cells onto a grid, a small flap of the formvar film was separated from the glass surface and floated over the grid. Subsequently, the preparation was drained of H2O on filter paper and allowed to dry. Cells that resided in open areas between the wires of the grid were available for examination.

High and intermediate voltage electron microscopes can be used to get good images of many types of thick biological specimens, including sections of embedded material up to several micrometers thick, whole cells, or isolated cellular components mounted on support films, stereoscopic pairs of images can be obtained by tilting the specimen between exposures, and this requires only relatively small tilt angles, often less than 15°. With a high tilt specimen holder, images covering a wider range of tilt angles can be obtained, increasing the amount of 3D information. Computer graphic methods for extracting quantitative 3D information from such images have been under development and use in this laboratory for several years.Many cellular structures take the form of networks. Examples that we study are transverse tubular networks in muscle, fibrin clots, and the microtrabecular lattice of the cytoplasm. We perform quantitative, 3D reconstruction and analysis of such networks from multiple stereoscopic image sets obtained using a JEOL 4000-EX TEM.

We have described the structure of virus assembly sites in frog virus 3-infected tissue culture cells based on an examination of sectioned and whole cells by conventional and high voltage (1 000 kV) electron microscopy (HVEM), respectively. We have also attempted to identify the cellular and viral components within the assembly sites using immunofluorescence and a combination of DNase digestion and EM autoradiography. Immunofluorescence studies showed that the sites do not contain tubulin, vimentin, actin, or myosin, confirming EM and HVEM studies which showed the absence of these cytoskeletal filaments from these sites. Enzymatic digestion and autoradiographic studies identified viral DNA in clumps of electron-dense material which are suspended within the matrix of the assembly site. Examination of whole, virus-infected cells by HVEM revealed that the matrix of the assembly site appears as a compaction of the cytoplasmic matrix and appears to be continuous with it. The fine strands of cytomatrix, equated with the microtrabeculae of other cells, are covered with or contain granules measuring 6–8 nm; similar granules also populate the virus assembly sites. Disruption of purified FV3 with chloroform yielded capsomeres measuring 6–8 nm. Based on all these observations, we postulate that the virus assembly sites are regions of the cytomatrix specialized for virus assembly and that the viral components are transported along the cytomatrix to the assembly sites.

The Holocentrus erythrophore, a red pigment cell, represents a model system for the study of organized intracellular transport. We have investigated the possibility that microtubules and actin are integral components of the pigment translocating motility machine. By creating cells that have total or partial loss of the microtubule framework we have demonstrated that the presence of microtubules is essential for organized, radial transport of the pigment granules. However, in the absence of microtubules, some undirected movement of the pigment can be stimulated; this suggests that a nonmicrotubular component of the cytoplast is responsible, at least in part, for the generation of motive force. In order to test the hypothesis that this component consists of actin or actomyosin, we examined the effects of probes for these classical motility proteins. Neither microinjection of phalloidin, DNase I or N-ethylmaleimide-modified heavy meromyosin nor exogenous application of cytochalasin B has any effect on pigment motion, although these materials do block the actin-mediated motility of other systems in our hands. Therefore, intracellular particle transport in erythrophores does not appear to be actin or actomyosin-based.

Most mammalian cells when placed in tissue culture are capable of autonomous motility, but the mechanisms and properties of this locomotion are not well understood. To test the behavior of cultured cells on substrata of variable composition, we used ordinary electron microscope grids as stencils while evaporating gold wire onto glass coverslips coated with cellulose acetate. Whole CHO or L-929 cells growing in monolayer culture were then trypsinized and plated on these prepared coverslips and fixed at various intervals for observation in the scanning electron microscope. In all cases studied, the whole cells were observed to accumulate on the gold areas in preference to the cellulose acetate. In addition, by centrifuging cells in the presence of cytochalasin B (CB), we obtained populations of enucleated cells (cytoplasts) which were then replated on these coverslips. Even though replated cytoplasts are capable of limited motility, they were not observed to preferentially select substrates as did intact cells.

We followed the translocation of identifiable pigment granules in living erythrophores through normal aggregation and dispersion and observed that they always return in dispersion to the same location relative to the whole pigment complex. This is interpreted to mean that each granule occupies a fixed position within a unit structure, the cytoplast. This position is retained even though the cytoplast undergoes dramatic reversals in form from ellipsoid to spheroid and back again with each aggregation and dispersion. The major structural components of the cytoplast, besides pigment granules, are microtubules and microtrabeculae. The latter constitute an irregular lattice that is confluent with microtubules and contains the pigment granules. In aggregation, the microtrabeculae shorten and seemingly contribute to the contraction of the entire cytoplast plus pigment. In dispersion, the microtrabeculae elongate in an apparent restructuring of the ellipsoidal cytoplast. The microtubules, however, persist in the cell cortex and appear to give radial direction to the pigment motion.

The cellular mechanisms used to direct translocating organelles are poorly understood. It is believed that the intrinsic structural polarity of microtubules may play a role in this process. We have examined the effects that differently oriented microtubules have upon the direction of pigment transport in surgically severed melanophore arms. In a previous paper (McNiven, M. A., M. Wang, and K. R. Porter, 1984, Cell, 37:753-765) we reported that after isolation, arms repolarized and reoriented their microtubules outward from their centers as if to form new "microcells." Pigment aggregation in these arms was toward a new focal point located at the arm centers. In this study we monitored pigment movement in isolated arms containing taxol-stabilized microtubules to test if the reversal in direction of pigment transport is dependent upon the repolarization of microtubules. We report that taxol delays both the microtubule reorientation and reversal in transport direction in a concentration-dependent manner. These and other presented data suggest that the polarity of the microtubule population within a melanophore confers direction on pigment transport.

The transport of pigment through the long cytoplasmic extensions (arms) of teleost melanophores is a microtubule-dependent event. We have severed the arms from melanophores to test whether microtubules isolated from the centrosome maintain their original polarity and disposition. In addition, we have tested whether arms containing microtubules of mixed polarities alter the direction of pigment transport. We find that microtubules within severed arms eventually change their polarity and reorganize from the arm center as if to form a new minicell. Concomitant with this change is a reversal in the direction of pigment transport.

WI-38 cells of intermediate and late-passage cultures were examined by light microscopy, scanning, conventional transmission and high voltage electron microscopy for evidence of heterogeneity among the cells within a single culture. Two morphologically distinct cell sizes and shapes were noted in all passages, (1) a typical fibroblastic type, and (2) a much larger, non-fusiform type. The larger cells generally had a nucleus that was positioned to one side of the bulk of the cytoplasm. The smaller was consistently fusiform with a centrally placed nucleus. The surfaces of intermediate-passage cells were uniform in showing small microvilli and corticall pits but were otherwise smooth and with out distinctive featues. The late-passage cells, on the other band, were consistent in showing numerous blebs and marginal ruffing. The internal structure of these cells in all passages studied was complicated by many age-related changes. The observations indicate that there is in WI-38 cells a high degree of intraculture heterogeneity. An awareness of this is important in studies which characterize the biochemical properties of this cell strain.

The integumental pigment cells (erythrophores) of the squirrel fish, Holocentrus ascensionis, are specialized for rapid radial transport of the pigment granules contained within their cytoplasm. Pigment granules in isolated denervated erythrophores alternate spontaneously between a centrally aggregated state and a radially dispersed state. In the absence of external calcium, pigment aggregation does not occur spontaneously and cannot be induced by the aggregating agents epinephrine or high concentration of external K + . Pigment aggregation is also impaired in the presence of D600 or papaverine, compounds reported to antagonize calcium influx into the cell. Pigment aggregation can be induced by experimental elevation of the concentration of cytoplasmic free Ca 2+ , with a Ca-EGTA buffer system in conjunction with ionophore A23187. The threshold concentration of Ca 2+ required to produce this effect is 5 × 10 −5 M. These results suggest that cytoplasmic free Ca 2+ is involved in mediating pigment aggregation and that some, if not all, the Ca 2+ is supplied by influx from the extracellular space.

Cytoskeletons with three-dimensional architecture were isolated from cultured normal rat kidney cells. The preparation procedure consisted of Triton-demembranization of suspended cells followed by differential and sucrose density gradient centrifugation. By using higher (0.5%) and lower (0.1%) Triton concentrations for demembranization, two kinds of isolated cytoskeletons (CSK), called H-CSK and L-CSK, respectively, were prepared. H-CSK and L-CSK displayed unique morphology and protein composition. Three classes of cytoskeletal filaments, microfilaments, intermediate filaments, and microtubules were shown to be major components in the electron microscopic images of the H-CSK. Stereoscopic electron microscopy of the H-CSK, dried by the critical point method, revealed that the cytoskeletal filaments are arranged in three-dimensional configurations even after isolation in vitro. Two-dimensional gel electrophoresis demonstrated that the H-CSK was composed mainly of actin, tubulin, and vimentin, reflecting its basic architecture. Electron microscopic images of L-CSK were more intricate than images of the H-CSK and showed, in addition to the filament types discussed above, anastomosing networks of short filamentous structures. These short filaments, with diameters of 3-8 nm and lengths of 30-150 nm, seemed to cross-link other elements of the cytoskeleton. The morphology of these short filaments resembles that of microtrabeculae observed in situ. Two-dimensional gels of the L-CSK showed over 100 protein spots when the gels were stained by the silver method. Subsequent treatment of the L-CSK with 0.5% Triton removed the microtrabeculae-like materials leaving as a residue the basic cytoskeleton similar to the H-CSK. Our observations indicate that microtrabeculae are composed of heterogenous proteins associated, in some instances, with a core structure of actin.

This paper presents the results of light and electron microscopy done on iridophores in the dorsal skin of the lizard Anolis carolinensis. New fine-structural details are revealed, and their importance is discussed. Of some interest is the complex of filaments between crystalline sheets in the cell. It is proposed that this complex is involved in the arrangement of crystals into crystalline sheets, and that the crystal arrangement and spacing are critical for the production of the cells' blue-green color. Tyndall scattering and thin-film interference are discussed as possible explanations for iridophore color production in relation to the fine-structural data obtained.

The serosal surfaces of major organs and tissues of the body cavities have been examined by scanning electron microscopy following fixation and critical point drying. The micrographs show that these surfaces are covered with microvilli and that the population is most dense on tissues that move about most actively in the body cavities. Radiating from the sides of the microvilli are many fine strands (150 A, average diameter) which interconnect with each other and adjacent microvilli. Transmission electron microscopic studies of material treated with thorium dioxide at low pH or stained with ruthenium red, reveal these strands to be (1) polyanionic in nature and (2) to arise from a surface glycocalyx which coats the microvilli and intervillar surfaces. It is proposed that the serosal microvilli, together with their glycocalyx, harbor a layer of serious exudate and thereby create a slippery cushion designed to protect the thin mesothelium from frictional damage.

SummaryProcedures are described for the successful culture of tissue cells in clots formed from purified bovine fibrinogen and thrombin. The influence on clot characteristics of pH and temperature, effective during clot formation, and the influence of fibrinogen concentration on the subsequent cell migration in the clot are indicated. Cultures grown in these clots can be transferred with the same ease as those grown in clots derived from chick plasma. The problem of fibrinoly-sis is briefly mentioned.

The observed sequences in the formation of clots from purified bovine fibrinogen and thrombin are described. Under the conditions of these experiments, it appears that fibrinogen molecules are polymerized by the action of thrombin to form needle-shaped, crystal-like protofibrils which then become aligned into fiber strands by lateral association. The integrity of the unit fibrils is maintained within the strand. A model of the fibrinogen molecule is proposed which may satisfy the reported physical constants, data from x-ray diffraction studies, and observations made upon electron micrographs.

The structure of the Z disc has been studied in thin sections of striated muscle fibers from a wide variety of vertebrates. A common organization is found in all muscles examined. The disc shows a regular pattern made up of dense lines which seem to connect the actin filaments from adjacent sarcomeres. The lines are sometimes disposed to form a regular zigzag configuration; in other orientations with respect to the plane of the section the morphology is confused and, in still other images, the dense lines continuous with the actin filaments seem to go straight through the Z disc. In cross section this structure corresponds to a square pattern of considerable regularity. The intersections in the square pattern mark the location in the plane of the section of the actin filaments from adjacent sarcomeres. Dense lines form the edges of the squares and appear to represent condensations of Z-disc material, i.e., the lines in the zigzag. The possible origin of the structure as a product of the stretching of a membrane is discussed, together with functional interpretations of the Z disc.

The electron microscope was used to study the structure and three dimensional relationships of the components of the body cortex in thin sections of Paramecium multimicronucleatum. Micrographs of sections show that the cortex is covered externally by two closely apposed membranes (together ∼250 A thick) constituting the pellicle. Beneath the pellicle the surface of the animal is molded into ridges that form a polygonal ridgework with depressed centers. It is these ridges that give the surface of the organism its characteristic configuration and correspond to the outer fibrillar system of the light microscope image. The outer ends of the trichocysts with their hood-shaped caps are located in the centers of the anterior and posterior ridges of each polygon. The cilia extend singly from the depressed centers of the surface polygons. Each cilium shows two axial filaments with 9 peripheral and parallel filaments embedded in a matrix and the whole surrouned by a thin ciliary membrane. The 9 peripheral filaments are double and these are evenly spaced in a circle around the central pair. The ciliary membrane is continuous with the outer member of the pellicular membrane, whereas the plasma membrane is continuous with the inner member of the pellicular membrane. At the level of the plasma membrane the proximal end of the cilium is continuous with its tube-shaped basal body or kinetosome. The peripheral filaments of the cilium, together with the material of cortical matrix which tends to condense around them, form the sheath of the basal body. The kinetodesma connecting the ciliary kinetosomes (inner fibrillar system of the light microscopist) is composed of a number of discrete fibrils which overlap in a shingle-like fashion. Each striated kinetosomal fibril originates from a ciliary kinetosome and runs parallel to other kinetosomal fibrils arising from posterior kinetosomes of a particular meridional array. Sections at the level of the ciliary kinetosomes reveal an additional fiber system, the infraciliary lattice system, which is separate and distinct from the kinetodesmal system. This system consists of a fibrous network of irregular polygons and runs roughly parallel to the surface of the animal. Mitochondria have a fine structure similar in general features to that described for a number of mammalian cell types, but different in certain details. The structures corresponding to cristae mitochondriales appear as finger-like projections or microvilli extending into the matrix of the organelle from the inner membrane of the paired mitochondrial membrane. The cortical cytoplasm contains also a particulate component and a system of vesicles respectively comparable to the nucleoprotein particles and to the endoplasmic reticulum described in various metazoan cell types. An accessory kinetosome has been observed in oblique sections of a number of non-dividing specimens slightly removed from the ciliary kinetosome and on the same meridional line as the cilia and trichocysts. Its position corresponds to the location of the kinetosome of the newly formed cilium in animals selected as being in the approaching fission stage of the life cycle.

An adepithelial basement lamella of 8–10 layers of apparently orthogonally arrayed collagen fibrils develops in the skin of Fundulus heteroclitus between the fifth and twelfth days after fertilization. Subsequent thickening of the lamella with growth until posthatching stages is the result of an increase in the number of rows of fibrils per layer, not in the number of layers. In contrast to the admuscular lamella, which forms in posthatching stages, the adepithelial lamella is acellular except in the sublamellar region. A morphological study of the skin of developing stages of Fundulus places the present “plywood” theory of basement lamella structure and development in doubt and supports a new “scindulene (shingle)” theory. Collagen fibril layers are observed to descend at a slight angle from the basement membrane, rather than to lie parallel to it as proposed by the plywood theory. Deposition of collagen in the basement lamella appears not to result from sequential switching and layering of collagen fibrils parallel to the basement membrane, as suggested by the plywood theory. Rather it appears that the collagen fibrils are inserted at end points into the basement membrane and are polymerized in one orientation only in any given area of the basement membrane and that adjacent areas are oriented at 105–110° to each other. Morphological changes with growth are accounted for by expansion of these “areas of insertion” in the basement membrane. Problems of variation in lamella thickness, crossover fibrils between different collagen layers, and the possible role of the basement membrane in the deposition of collagen into the lamellar structure are discussed.

Prior to gastrulation, the microtubules in the presumptive primary mesenchyme cells appear to diverge from points (satellites) in close association with the basal body of the cilium; from here most of the microtubules extend basally down the lateral margins of the cell. As these cells begin their migration into the blastocoel, they lose their cilia and adopt a spherical form. At the center of these newly formed mesenchyme cells is a centriole on which the microtubules directly converge and from which they radiate in all directions. Later these same cells develop slender pseudopodia containing large numbers of microtubules; the pseudopodia come into contact and fuse to form a "cable" of cytoplasm. Microtubules are now distributed parallel to the long axis of the cable and parallel to the stalks which connect the cell bodies of the mesenchyme cells to the cable. Microtubules are no longer connected to the centrioles in the cell bodies. On the basis of these observations we suggest that microtubules are a morphological expression of a framework which opeartes to shape cells. Since at each stage in the developmental sequence microtubules appear to originate (or insert) on different sites in the cytoplasm, the possibility is discussed that these sites may ultimately control the distribution of the microtubules and thus the developmental sequence of form changes.

Pigment epithelial cells of the frog's retina have been examined by methods of electron microscopy with special attention focused on the fine structure of the endoplasmic reticulum and the myeloid bodies. These cells, as reported previously, send apical prolongations into the spaces between the rod outer segments, and within these extensions, pigment migrates in response to light stimulation. The cytoplasm of these cells is filled with a compact lattice of membrane-limited tubules, the surfaces of which are smooth or particle-free. In this respect, the endoplasmic reticulum here resembles that encountered in cells which produce lipid-rich secretions. The myeloid bodies comprise paired membranes arranged in stacks shaped like biconvex lenses. At their margins the membranes are continuous with elements of the ER and in consequence of this the myeloid body is referred to as a differentiation of the reticulum. The paired membranes resemble in their thickness and spacings those which make up the outer segments; they are therefore regarded as intracellular photoreceptors of possible significance in the activation of pigment migration and other physiologic functions of these cells. The fuscin granules are enclosed in membranes which are also continuous with those of the ER. The granules seem to move independently of the prolongations in which they are contained. The report also describes the fine structure of the terminal bar apparatus, the fibrous layer intervening between the epithelium and the choroid blood vessels, and comments on the functions of the organelles depicted.

This paper reports on the fine structure of rat oocytes at stages before ovulation, during maturation, fertilization, and early cleavage. The study includes parallel observations on light and electron microscope preparations with attempted correlations. The follicular cells of the ovarian egg are described as sending long processes through the zona pellucida to the egg surface where they mingle with thin projections from the egg itself. No open communication between follicle cell cytoplasm and egg cytoplasm was observed. During maturation and fertilization both types of processes are withdrawn from the zona. The germinal vesicle and later the pronuclei of the fertilized egg are characterized by numerous large nucleoli. These have the form of thick walled vesicles with diameters as great as 8 to 10 µ. The wall is dense in the EM image and appears to consist in part of small granules. The cytoplasm shows several inclusions including mitochondria of usual form and a Golgi component which has the typical fine structure and the distribution described by earlier light studies. Small dense particles, presumably RNP particles, are distributed throughout the cytoplasmic matrix and show no preference for membranes. The endoplasmic reticulum of the oocyte is represented by a scattering only of vesicles, but begins a more extensive and elaborate development with the onset of segmentation. One inclusion of the ooplasm, similar in size to mitochondria, receives special attention. It is a vesicular structure, containing a large number of small vesicles (10 to 50 mµ in diameter) and frequently a central density or nucleoid. They are referred to as multivesicular bodies. Such bodies are found in small number in the ovarian egg, but increase greatly in number during maturation and fertilization. It appears from the micrographs of eggs in these latter stages that these vesicular bodies break down and liberate their content of small vesicles to the surrounding ooplasm. Comments are provided on the apparent significance of the various observations.

Spermiogenesis in chicken has been examined in order to see whether the radical changes observed in cell shape can be related to the presence of cytoplasmic microtubules. A highly ordered array of tubules has been found which surrounds the nucleus as it elongates from a sphere to a slender cylinder. The structure of the array has been determined by following the tubules through 12–14 adjacent serial sections, and it is a left-handed double helix. Faint cross-bridges connect consecutive turns of the two helices. After the change in nuclear shape is complete, the helical system of microtubules disappears and is replaced by a set of almost straight tubules which run parallel to the long axis of the nucleus. These tubules remain while the spermatid nucleus condenses isotropically to its final size. We suggest that the helix is the agent which effects nuclear elongation and that the subsequent system of paraxial tubules determines the curvature of the final sperm head. Evidence for these suggestions is found in the form of spermatids which have failed to develop properly. In an appendix we consider the kinematics of single and multiple helix systems and discuss the revelance of these models to the morphogenesis of chicken spermatids.

Skeletal myofibrils isolated either by tryptic digestion at 0° C. or by a colloid mill and suspended in buffer solution (pH 7.0, µ; 0.154) containing 20 per cent glycerin and 0.0025 M adenosinetriphosphate at –5° C. contracted slowly and progressively when the temperature was raised above 0° C. Formalin fixation halted this contraction. With the aid of these procedures myofibrils in progressive stages of contraction were then studied with the electron microscope. Electron micrographs showed that uncontracted fibrils isolated by the colloid mill were structurally similar to those described by other workers. Treatment of fibrils with trypsin removed the Z bands and disorganized the I bands. This enzymatic modification of structure did not impair the contractile response. The principal structural changes during contraction consisted of a migration of dense material from the A band into the A-I junction or the Z band, a gradual increase in width of the fibril, a gradual decrease in length of sarcomeres, an apparent increase in the mean diameter of filaments, and a disorientation of these latter from their parallel arrangement.

This report provides information on the morphology of fat absorption in rat intestinal epithelial cells. Three types of experiments were performed: (a) intubation of corn oil into fasted rats, (b) injection of physiological fatty-chyme prepared from fat-fed donor rats into ligated segments of jejunum of fasted animals, and (c) administration of electron-opaque particles in corn oil and markers given concurrently with the fat. These results support the hypothesis that fat is absorbed by selective diffusion of monoglycerides and fatty acids from micelles rather than by pinocytosis of unhydrolized triglycerides. Evidence is presented that the pits between the microvilli, previously believed to function in the transport of fat, are not involved in this process. Instead they appear to contribute their contents to lysosomes in the apical cytoplasm. Arguments are offered that the monoglycerides and fatty acids diffuse from the micelle while the latter is associated with the microvillous membrane of the absorptive cell. These micellar components penetrate the plasma membrane and diffuse into the cytoplasmic matrix where they encounter the SER. Triglyceride synthesis occurs in the SER and results in the deposition of fat droplets within its lumina. The synthesis of triglycerides and their sequestration into the SER establishes an inward diffusion gradient of monoglycerides and fatty acids.

Cells of onion and garlic root tips were examined under the electron and phase contrast microscopes after fixation in KMnO4. Special attention was focused on the distribution and behavior of the endoplasmic reticulum (ER) during the several phases of mitosis. Slender profiles, recognized as sections through thin lamellar units of the ER (most prominent in KMnO4-fixed material), are distributed more or less uniformly in the cytoplasm of interphase cells and show occasional continuity with the nuclear envelope. In late prophase the nuclear envelope breaks down and its remnants plus cytoplasmic elements of the ER, which are morphologically identical, surround the spindle in a zone from which mitochondria, etc., are excluded. During metaphase these ER elements persist and concentrate as two separate systems in the polar caps or zones of the spindle. At about this same time they begin to proliferate and to invade the ends of the spindle. The invading lamellar units form drape-like partitions between the anaphase chromosomes. In late anaphase, their advancing margins reach the middle zone of the spindle and begin to fray out. Finally, in telophase, while elements of the ER in the poles of the spindle coalesce around the chromosomes to form the new envelope, the advancing edges of those in the middle zone reticulate at the level of the equator to form a close lattice of tubular elements. Within this, which is identified as the phragmoplast, the earliest signs of the cell plate appear in the form of small vesicles. These subsequently grow and fuse to complete the separation of the two protoplasts. Other morphological units apparently participating in mitosis are described. Speculation is provided on the equal division or not of the nuclear envelope and the contribution the envelope fragments make to the ER of the new cell.

By means of a tissue culture technique, cells from chick embryos were procured in a state which proved to be suitable for electron microscopy. The electron micrographs disclosed details of cell structure not revealed by other methods of examination.

The cytoplasmic ground substance of animal tissue cells grown in vitro has been found by electron microscopy to contain, as a part of its submicroscopic structure, a complex reticulum of strands, to be referred to as the endoplasmic reticulum. It has been found in all types of cells extensively studied. The components of this reticular system vary considerably in size and form, apparently in some relation to physiological changes in the cell. Thus in one cell of a culture colony it may be finely divided into strands or canaliculi, 50 to 100 mµ in diameter, whereas in an adjacent cell of the same type the components of the reticulum may be relatively coarse, 600 mµ in diameter, and vesiculated. The membrane, which can be shown to limit the system and separate it from the rest of the ground substance, is similar in thickness to the plasma membrane surrounding the cell. Photomicrographs of living cells taken by phase contrast and dark field microscopy define a structure of similar form and indicate that the reticulum of the electron microscope image has its equivalent in the living unit. Where its component units are sufficiently large, a structure of identical form can be resolved by light microscopy in cells stained with hematoxylin or with toluidine blue. This indicated that the endoplasmic reticulum is to be identified with the basophilic or chromophilic component (the ergastoplasm) of the cytoplasm and that such properties of this component as have been determined by cytochemical methods, such as a high RNA content, may be assigned to this "submicroscopic" system.

Cells of the mouse line Balb/3T3 as well as three virus-induced transformants and two spontaneous transformants grown in vitro have been studied for their topography by scanning electron microscopy. The parent cell in confluent culture closely resembles an endothelial cell in its form and in the structure of its association with adjacent cells. The tumorigenic transformants produced by SV40, murine sarcoma virus, or polyoma viruses are fusiform to pleomorphic and distinctly different from the cell of origin. They show relatively smooth surfaces except for blebs and marginal microvilli. Perhaps most surprising is the similarity they bear to one another. This is made the more singular by the very different form shown by the tumorigenic transformants of spontaneous origin. One of these, S2-4, possesses a thickened rather than the lamellar form of the parent A31 cell and is covered by long microvilli and many spherical blebs. The other, TuT3, more closely resembles the cell of origin but shows extensive ruffling at its margins. All transformants grow without evidence of contact inhibition. The significance of the surface morphologies and the factors influencing cell form are discussed.