UNUSUAL MICROTUBULAR PATTERNS AND THREE-DIMENSIONAL MOVEMENT OF MEALYBUG SPERM AND SPERM BUNDLES

The spermatozoon of the mealybug Pseudococcus obscurus Essig is a filamentous cell (0.25 µ by 300 µ) which exhibits three-dimensional flagellations throughout most of its length. It has microtubules (200 A diameter) and a threadlike nuclear core (0.07–0.09 µ diameter) which extend almost its entire length, but apparently it has no mitochondria, centrioles, typical flagellum, or acrosome. The microtubules are arranged in two and a half concentric rings and total 56 in the most actively motile region but form two or three concentric rings with totals of 28 or 56 tubules, respectively, in less active regions. The relation of unusual microtubular patterns to the 9 + 2 complex and to flagellar motion is discussed. Mealybug spermatozoa are transmitted to the female in motile bundles which are approximately 1.3 µ by 750 µ and have four regions: (1) an anterior corkscrew region; (2) a region which contains approximately 16 spermatozoa; (3) a region of amorphous content; and (4) an endpiece. Bundle motility originates from the synchronous movements of its spermatozoa which appear to be arranged in two concentric multistranded helices. The spermatozoa provide both forward and gyratory motions of the bundle, and the corkscrew complements bundle propulsion by converting part of the rotation into forward movement.     

The fine structure of the sperm bundles of the coccid,Aspidiotus perniciosus Cumst., and sperm movement

The mature sperm of A. perniciosus are organized into bundles, about 350 μm long by 9–10 μm wide. Each bundle contains 32 sperm enclosed by a common sheath. The sperm contains an elongated ‘central core’, representing nuclear material, surrounded by a spiral microtubular sheath and cytoplasm. The electron-dense nuclear material is localized in the more pointed half of the sperm. The spiral microtubular sheath is composed of 30— 100 microtubules (depending on the cross-sectional level), situated parallel to the longitudinal axis of the sperm. On the basis of this ultrastructural organization, the motility of the sperm and sperm bundle as a whole is discussed. The sperm of A. perniciosus provide strong evidence that the microtubules arranged asymmetrically represent the elements directly involved in sperm motility.     

MICROTUBULAR PATTERNS IN SPERMATOZOA OF COCCID INSECTS IN RELATION TO BENDING

Flagella-like motion occurs in filamentous spermatozoa of coccid insects, which have diameters (0.16–0.65 µ) and lengths (150–300 µ) similar to those of long flagella, but have no doublets or 9 + 2-like arrangements of microtubules. Light and electron microscope investigations of spermatozoa from 10 species reveal many bizarre patterns of microtubules and suggest some basic similarities to flagella. Detailed analyses of spermatozoa which are naturally bent in definable planes during their elongation in the male and their storage in the female provide evidence that a constant topographical relationship is maintained between their unorthodox patterns of microtubules, as viewed in transections, and the direction of bending. The configuration common to most coccid spermatozoa consists of an acentrically positioned crescent of microtubules surrounded by one to several concentric rings. A line drawn to connect the two ends of the crescent appears to remain perpendicular to the plane of bending, and it defines a plane in which bisection of the spermatozoon produces halves with unequal numbers of microtubules. Bisection of the 9 + 2 motile apparatus in a plane perpendicular to that of bending also appears to produce halves with unequal numbers of microtubules. Therefore, the indispensable elements for flagellar and flagella-like motion may be microtubules arranged in "asymmetric" patterns.     

ON THE ROLE OF MICROTUBULES IN MOVEMENT AND ALIGNMENT OF NUCLEI IN VIRUS-INDUCED SYNCYTIA

Infection of baby hamster kidney (BHK21-F) cells with the parainfluenza virus SV5 causes rapid and extensive cell fusion. Time-lapse cinematography shows that when cells fuse, their nuclei migrate straight to the center of the syncytium at rates of 1–2 µ/min. Nuclei are often arranged in long, tightly packed, parallel rows in syncytia derived from the fibroblastic BHK21-F cells. Polarization microscopy shows birefringent material between and parallel to these rows of nuclei, and electron microscopy shows bundles of cytoplasmic microtubules, ~250 A in diameter, and filaments, ~80 A in diameter, parallel to and between the rows of nuclei. Colchicine treatment causes disappearance of microtubules from BHK21-F cells and an apparent increase in the number of 80-A filaments. Although colchicine-treated, SV5-infected cells fuse, their nuclei do not migrate or form rows but remain randomly scattered through the syncytial cytoplasm. Incubation at 4°C does not disrupt microtubules in BHK21-F cells. Rows of nuclei have been isolated from SV5-induced syncytia, and the nuclei in them have been found to be intimately associated with microtubules but not with other cytoplasmic structures. These results suggest that microtubules demarcate cytoplasmic channels through which nuclei migrate and that they may also be involved in the mechanism of nuclear movement.     

Ultrastructure of the spermatozoon of Anomotaenia quelea (Mettrick, 1961) (Cestoda,Cyclophyllidea, Dilepididae), an intestinal parasite of Quelea quelea (Aves,Ploceidae) in Senegal

The mature spermatozoon of Anomotaenia quelea exhibits an apical cone of electron-dense material and two helicoidal crest-like bodies. The apical cone near its base is surrounded by a lucent cytoplasm and a spiraled layer of cortical microtubules. The crest-like bodies are of different lengths, spiraled and make an angle of 30–40° to the hypothetical spermatozoon axis. The axoneme is of the 9 + ‘1’ trepaxonematan pattern and is surrounded by a periaxonemal sheath of electron-dense material. The cytoplasm contains in regions III and IV numerous electron-dense granules situated between the periaxonemal sheath and the cortical microtubules. The posterior extremity of the spermatozoon of A. quelea exhibits a nucleus and a disorganized axoneme and cortical microtubules. This type of posterior extremity of the mature spermatozoon has never been described previously in a Dilepididae. Similarly, two crest-like bodies have not been observed before in a dilepidid cestode.     

THE STRUCTURAL BASIS OF CILIARY BEND FORMATION : Radial Spoke Positional Changes Accompanying Microtubule Sliding

The sliding microtubule model of ciliary motility predicts that cumulative local displacement (Δl) of doublet microtubules relative to one another occurs only in bent regions of the axoneme. We have now tested this prediction by using the radial spokes which join the A subfiber of each doublet to the central sheath as markers of microtubule alignment to measure sliding displacements directly. Gill cilia from the mussel Elliptio complanatus have radial spokes lying in groups of three which repeat at 860 Å along the A subfiber. The spokes are aligned with the two rows of projections along each of the central microtubules that form the central sheath. The projections repeat at 143 Å and form a vernier with the radial spokes in the precise ratio of 6 projection repeats to 1 spoke group repeat. In straight regions of the axoneme, either proximal or distal to a bend, the relative position of spoke groups between any two doublets remains constant for the length of that region. However, in bent regions, the position of spoke groups changes systematically so that Δl (doublet 1 vs. 5) can be seen to accumulate at a maximum of 122 Å per successive 860-Å spoke repeat. Local contraction of microtubules is absent. In straight regions of the axoneme, the radial spokes lie in either of two basic configurations: (a) the parallel configuration where spokes 1–3 of each group are normal (90°) to subfiber A, and (b) the tilted spoke 3 configuration where spoke 3 forms an angle (θ) of 9–20°. Since considerable sliding of doublets relative to the central sheath (~650 Å) has usually occurred in these regions, the spokes must be considered, functionally, as detached from the sheath projections. In bent regions of the axoneme, two additional spoke configurations occur where all three spokes of each group are tilted to a maximum of ± 33° from normal. Since the spoke angles do not lie on radii through the center of bend curvature, and Δl accumulates in the bend, the spokes must be considered as attached to the sheath when bending occurs. The observed radial spoke configurations strongly imply that there is a precise cycle of spoke detachment-reattachment to the central sheath which we conclude forms the main part of the mechanism converting active interdoublet sliding into local bending.     

Ultrastructure of the spermatozoon of Anomotaenia quelea (Mettrick, 1961) (Cestoda,Cyclophyllidea, Dilepididae), an intestinal parasite of Quelea quelea (Aves,Ploceidae) in Senegal

The mature spermatozoon of Anomotaenia quelea exhibits an apical cone of electron-dense material and two helicoidal crest-like bodies. The apical cone near its base is surrounded by a lucent cytoplasm and a spiraled layer of cortical microtubules. The crest-like bodies are of different lengths, spiraled and make an angle of 30–40° to the hypothetical spermatozoon axis. The axoneme is of the 9 + ‘1’ trepaxonematan pattern and is surrounded by a periaxonemal sheath of electron-dense material. The cytoplasm contains in regions III and IV numerous electron-dense granules situated between the periaxonemal sheath and the cortical microtubules. The posterior extremity of the spermatozoon of A. quelea exhibits a nucleus and a disorganized axoneme and cortical microtubules. This type of posterior extremity of the mature spermatozoon has never been described previously in a Dilepididae. Similarly, two crest-like bodies have not been observed before in a dilepidid cestode.     

CYTOPLASMIC MICROTUBULE AND WALL MICROFIBRIL ORIENTATION IN ROOT HAIRS OF RADISH

The fine structure of young root hairs of radish was studied, with special attention to cytoplasm-wall relationships. Hairs up to 130 µ in length were examined after fixation of root tips in glutaraldehyde followed by osmium tetroxide. Microtubules occur axially aligned in the cytoplasm just beneath the plasmalemma, and extend from the base of the hair to within 2 to 3 µ of the tip. Poststaining with uranyl acetate and lead citrate clearly reveals in thin sections the presence of the two layers of cellulose microfibrils known from studies on shadowed wall preparations: an outer layer of randomly arranged microfibrils arising at the tip, and a layer of axially oriented microfibrils deposited on the inside of this layer along the sides. The youngest microfibrils of the inner, oriented layer first appear at a distance of about 25 µ from the tip. Although the microfibrils of the inner layer and the adjacent microtubules are similarly oriented, the oriented microtubules also extend through the 20- to 25-µ zone near the tip where the wall structure consists of random microfibrils. This suggests that the role of microtubules in wall deposition or orientation may be indirect.     

SPERMIOGENESIS IN WILD TYPE AND IN A MALE STERILITY MUTANT OF DROSOPHILA MELANOGASTER

Spermiogenesis in the translocation heterozygote T (1; 2H) 25(20) y l 25/FM6 has been studied with the electron microscope and compared with that in wild type males. It appears that the genetic lesion in the male sterility mutant is associated primarily with a failure in differentiation of the head. In wild type flies, the spermatid nucleus assumes a conchoidal shape; chromatin accumulates along the convex surface. Adjacent to the concave surface a large bundle of microtubules runs parallel to the long axis of the spermatid. A single row of microtubules is juxtaposed against the convex surface of the head. As differentiation proceeds, the nucleus elongates, chromatin condenses, and the nucleus is compacted to a final diameter of about 0.3 µ. In the sterile mutant the spermatid nucleus has an irregular or wedge-shaped profile and no concavity is formed, nor is the bundle of microtubules observed. The row of microtubules, however, is usually present around the periphery. The change from lysine-rich to arginine-rich histone in mature wild type sperm does not occur in the sterile male. The substructure of the axial filament and mitochondrial derivatives, however, are similar to those in wild type.     

Spermiogenesis in an iceryine coccid,Steatococcus tuberculatus morrison

The spermatozoon of Steatococcus is a motile filament containing a core of two chromosomes arranged in tandem and surrounded by more than 80 microtubules in 2 1/2 concentric rings. Two sperm develop from each binucleate spermatid in the form of long papillae. From the zone corresponding to the pole of the previous division microtubules appear and lengthen, assembly apparently occurring at their proximal undifferentiated ends. As they extend, they presumably push out the cytoplasmic papilla and co-extend a nuclear papilla through bridges with the nuclear envelope. Chromatin, attached to the envelope, is thus carried into the papilla, the shorter chromosome in the lead. 100 Å chromatin filaments are reduced to 20 Å and aligned as they enter the papilla. The filaments transform into 100 Å tubular fibrils, presumably by supercoiling. These then pack hexagonally, aggregate further into packed axial filaments, and finally condense into a nearly solid core in the mature sperm. Completed papillae (sperm) detach from the spermatid leaving behind nuclei devoid of chromatin. Following cycles of spiralization and despiralization, the sperm are bundled into hexagonal packs of 32 in register by cyst wall cells. The latter form primary and secondary sheaths and lay down a matrix within the bundle. As originally reported by Hughes Schrader (1946), no evidence of centriole, acrosome, mitochondrial derivative or structure suggesting flagellar axoneme is found in either the developing papilla or the mature sperm. The microtubules determine the axis of the developing sperm; polarity is set by the direction of sperm motion and is homologous with most flagellate sperm in that the nuclear material is anterior and the microtubule initiating center is posterior. All of the functions attributed to microtubules are manifest in differentiation of this sperm: extension, support, translocation and motility.This paper is affectionately dedicated to Professor Sally Hughes-Schrader on the occasion of her seventy-fifth birthday, with warm appreciation of her friendship, her exemplary science, her keen criticism, her contagious enthusiasm, and for leading me to Steatococcus.     

PRODUCTION AND RELEASE OF A LOCUS-SPECIFIC RIBONUCLEOPROTEIN PRODUCT IN POLYTENE NUCLEI OF DROSOPHILA HYDEI

A specific, 0.1–0.3-µm large ribonucleoprotein complex consisting of a central core with stalklike extensions on top of which 280–320-Å ribonucleoprotein particles are situated is found in an experimentally activated chromosome region, 2–48C, of the polytene chromosomes of Drosophila hydei. Alkaline hydrolysis, RNAse digestion, and uranyl-EDTA-lead staining indicated the ribonucleoprotein character of the 280–320-Å particles, whereas the central core seems to be devoid of RNA. The characteristic complexes are present in the nucleoplasm and at the nuclear membrane, but absent from the cytoplasm. It is suggested that the large RNP complexes are the specific products of the puff at 2–48C. Complexes similar to the ones described have not been observed in any other region of the polytene salivary gland chromosomes of this species.     

ELECTRON MICROSCOPIC STUDY OF THE CYTOPATHOLOGY OF ECHO VIRUS INFECTION IN CULTIVATED CELLS

The cultivated monkey kidney cell is subject to changes when infected with ECHO viruses 6, 9, and 19. The electron microscope reveals three stages of infection: (a) initial stage. The nucleus appears granular with chromatin condensation on the nuclear envelope. The cytoplasm contains electron transparent vesicles and vacuoles forming nests. (b) Intermediate stage. The nucleus seems to diminish, appearing more pycnotic and displaced toward the periphery. The cytoplasm is filled with electron transparent vacuoles and vesicles, and dense masses as well as some spiral bodies are seen. The mitochondria retain their shape. Dense particles are seen, which are possibly of viral nature. (c) Final stage. The nucleus is contracted to a narrow strip close to the cellular membrane or is completely destroyed. The cytoplasm shows no apparent changes. Crystals are frequently observed in cells infected with ECHO viruses 6 and 19, consisting of dense particles with an average diameter of 14.4 mµ ranging from approximately 13.2 to 15.6 mµ for ECHO virus 6, and 14.5 mµ ranging from approximately 12.5 to 16.5 mµ for ECHO virus 19. These particles are clustered in hexagonal packages forming angles of 75° and 105°. The particles in most crystals are arranged in rows separated by a constant distance, the latter varying from one crystal to another and being approximately 1.5 and 2.5 times the distance between particles. Other particles were observed which, however, are not considered to be of viral nature.     

Microtubules and filaments in the axons and astrocytes of early postnatal rat optic nerves

Changes in the population of microtubules and filaments within the cytoplasm of maturing axons and astrocytes have been studied during the early postnatal development of rat optic nerves. At birth, all of the axons are unmyelinated; most have a diameter of 0.2–0.3 µ and contain 4–10 microtubules. Neurofilaments do not occur with any frequency until about 5 days postnatal when they appear as individual groups, each containing 4–12. Subsequently, the neurofilaments of each group disperse so that they become more evenly distributed in mature axons. Developing astrocytes show similar but rather more dramatic changes. Most astrocytic processes contain only microtubules at birth, but during maturation filaments begin to appear in increasing numbers while microtubules become less common. This process continues until, in the mature fibrous astrocytes, filaments pack the cytoplasm and microtubules are rare. These observations suggest that the filaments within axons and astrocytes may be formed by the breakdown of microtubules.     

THE AXOSTYLE OF SACCINOBACULUS : II. Motion of the Microtubule Bundle and a Structural Comparison of Straight and Bent Axostyles

Undulations of the flagellate Saccinobaculus result from motility in its axostyle, a bundle consisting of thousands of cross-bridged microtubules. In its resting state, the axostyle is a helix of large pitch and slowly varying radius. The active state as seen by light microscopy involves first a bending of the anterior end of the axostyle to a radius of about 8 µm with a circular arc ranging from 60° to 180°, and then the propagation of this bend without damping to the posterior end of the organism at speeds up to 100 µm/s. The cross section of an unbent axostyle is crescent shaped. This crescent flattens as the bend arrives and reappears as the bend passes by. Intertubule bridges impart to the axostyle tubules an axial periodicity of about 150 Å which can serve as a marker for the investigation of tubule sliding or contraction associated with bend formation. Optical diffraction measurements on electron micrographs of the bend demonstrate that the axostyle tubules slide over one another and that the tubules on the inside of a bend usually contract, sometimes by as much as 25%. Possible relationships between the contraction and sliding of the tubules are discussed.     

STRUCTURE AND DEVELOPMENT OF VIRUSES OBSERVED IN THE ELECTRON MICROSCOPE : IV. VIRUSES OF THE RI-APC GROUP

Representative viruses of the RI-APC group were observed with the electron microscope in thin sections of infected HeLa cells. The viral particles varied in density, were approximately 60 mµ in diameter and had a center to center spacing when close packed of about 65 mµ. Many of the less dense particles exhibited an internal body averaging 24 mµ in diameter. It was suggested that within the nucleus the virus differentiated from dense granular and reticular material and formed crystals. Disintegration of the crystals and disruption of the nuclear membrane with release of virus into the cytoplasm appeared to occur at any stage. No evidence to suggest development of the virus in the cytoplasm was obtained. It was possible to deduce the structure of the viral crystal from the electron micrographs. The viral particles are packed in a cubic body—centered lattice. Correlative histochemical observations in the light microscope which are now in progress revealed that the crystals and non-crystalline aggregates of virus were strongly Feulgen-positive.     

MICROTUBULES AND EARLY STAGES OF CELL-PLATE FORMATION IN THE ENDOSPERM OF HAEMANTHUS KATHERINAE BAKER

A fine structure study of the phragmoplast and developing cell plate has been made on glutaraldehyde-osmium tetroxide-fixed, dividing, cultured cells of the liquid endosperm of Haemanthus katherinae Baker. The phragmoplast arises between the telophase nuclei, usually in association with a remnant strand of spindle elements, and consists of an accumulation of microtubules oriented at right angles to the plane of the future cell plate. The microtubules, which are 200–240 A in diameter, occur in small clusters spaced at approximately 0.2–0.3 µ intervals along the plate. Short interconnections interpreted as "cross-bridges" have been observed between individual microtubules. Within each cluster there is an electron-opaque zone about 0.3 µ in width which can be attributed in part to an overlap of microtubules from both sides of the plate and in part to a local accumulation of an amorphous electron-opaque material. During development these dense zones become aligned in a plane which itself defines the plane of the plate. Vesicles, commonly observed in long files, are derived from a cytoplasmic matrix rich in elements of the endoplasmic reticulum and sparse in dictyosomes. They aggregate between the clusters of microtubules and eventually coalesce to form the cell plate.     

ELECTRON MICROSCOPY OF GROWING OOCYTES OF RANA PIPIENS

1. In the cytoplasm of oocytes of stage Y₀, prior to the appearance of yolk, one observes a few scattered profiles of endoplasmic reticulum and numerous filamentous mitochondria, usually distributed at random but sometimes clustered. As the nuclear membrane begins to bulge outward, small granules and short rods appear in the perinuclear cytoplasm and endoplasmic reticulum becomes more prominent throughout the cytoplasm. 2. Coincident with the appearance of the first yolk platelets, which are deposited in a narrow peripheral ring within the endoplasm at stage Y₁, protoplasmic processes, the microvilli, push out all over the surface of the oocyte. At the same time follicle cells pull away but remain attached to the oocyte at some points through finger-like processes which interdigitate with neighboring microvilli. It is estimated that the microvilli increase the absorptive area of the surface to about thirty-five times that of a simple sphere. Just beneath the microvillous layer is the basal protoplasm of the cortex, now containing tiny granules probably synthesized from newly absorbed raw materials. Cortical granules appear and become aligned below the basal layer on the external border of the endoplasm. Both the cortical granules and the yolk platelets measure up to 1 µ in diameter at this stage. 3. By stage Y₃ (yolk filling peripheral three-fourths of cytoplasm), the basal layer of the cortex is folded so that it appears in section as alternating ridges and valleys. The microvilli now extend from the summits of the cortical ridges. Small, ring-shaped granules are abundant in the cortex. Cortical granules have increased to 2 µ in diameter. 4. Yolk platelets continue to be synthesized around the cortical granules and in the subjacent endoplasm. The largest platelets measured in the interior cytoplasm at stage Y₄ (cytoplasm filled with yolk) were 3.7 µ wide by 5.8 µ long. Pigment granules increase in size from 0.15 µ in diameter at stage Y₃ to 0.30 µ in diameter at stage Y₄.     

First cytochemical study of haemocytes from the crab Carcinus aestuarii (Crustacea,Decapoda)

For the first time, a morphological study of haemocytes from the crab Carcinus aestuarii was carried out by means of light microscopy and differing cytochemical assays. Analysis of haemocyte size frequency distribution (performed by means of a Coulter Counter) revealed the presence of two distinct haemocyte fractions in C. aestuarii haemolymph, depending on cell size. The first fraction was of about 3–5 µm in diameter and 30–50 fL in volume, the second was of about 6–12 µm in diameter and over 200 fL in volume. Mean cell diameter and volume were 8.20±1.7 µm and 272.30±143.5 fL, respectively. Haemocytes observed under light microscope were distinguished in three cell types: granulocytes (28%; 11.94±1.43 µm in diameter) with evident cytoplasmic granules, semigranulocytes (27%; 12.38±1.76 µm in diameter) with less granules than granulocytes, and hyalinocytes (44%; 7.88±1.6 µm in diameter) without granules. In addition, a peculiar cell type was occasionally found (about 1%): it was 25–30 µm in diameter and had a great vacuole and a peripheral cytoplasm with granules. Granulocyte and semigranulocyte granules stained in vivo with Neutral Red, indicating that they were lysosomes. Giemsa’s dye confirmed that granulocytes and semigranulocytes were larger than hyalinocytes. Pappenheim’s panoptical staining and Ehrlich’s triacid mixture allowed to distinguish granule-containing cells (including semigranulocytes) in acidophils (64%), basophils (35%) and neutrophils (1%). Hyalinocytes showed always a basophilic cytoplasm. Haemocytes were positive to the PAS reaction for carbohydrates, even if cytoplasm carbohydrate distribution varied among cell types. Lastly, lipids were found on cell membrane and in cytoplasm of all haemocyte types in the form of black spots produced after Sudan Black B staining. The morphological characterisation of C. aestuarii haemocytes by light microscopy was necessary before performing both ultrastructural and functional studies of circulating cells.Key words: Carcinus aestuarii, crab, haemocytes, light microscopy, cytochemical assays, morphological characterisation.     

THE FINE STRUCTURE OF COCKROACH CAMPANIFORM SENSILLA

Campaniform sensilla on cockroach legs provide a good model system for the study of mechanoreceptive sensory transduction. This paper describes the structure of campaniform sensilla on the cockroach tibia as revealed by light- and electron-microscopy. Campaniform sensilla are proprioceptive mechanoreceptors associated with the exoskeleton. The function of each sensillum centers around a single primary sense cell, a large bipolar neuron whose 40 µ-wide cell body is available for electrophysiological investigation with intracellular microelectrodes. Its axon travels to the central nervous system; its dendrite gives rise to a modified cilium which is associated with the cuticle. The tip of the 20 µ-long dendrite contains a basal body, from which arises a 9 + 0 connecting cilium. This cilium passes through a canal in the cuticle, and expands in diameter to become the sensory process, a membrane-limited bundle of 350–1000 parallel microtubules. The tip of the sensory process is firmly attached to a thin cap of exocuticle; mechanical depression of this cap, which probably occurs during walking movements, effectively stimulates the sensillum. The hypothesis is presented that the microtubules of the sensory process play an important role in mechanoelectric transduction in cockroach campaniform sensilla.     

STUDIES ON THE MICROTUBULES IN HELIOZOA : II. The Effect of Low Temperature on These Structures in the Formation and Maintenance of the Axopodia

When specimens of Actinosphaerium nucleofilum are placed at 4°C, the axopodia retract and the birefringent core (axoneme) of each axopodium disappears. In fixed specimens, it has been shown that this structure consists of a highly patterned bundle of microtubules, each 220 A in diameter; during cold treatment these microtubules disappear and do not reform until the organisms are removed to room temperature. Within a few minutes after returning the specimens to room temperature, the axonemes reappear and the axopodia begin to reform reaching normal length 30–45 min later. In thin sections of cells fixed during the early stages of this recovery period, microtubules, organized in the pattern of the untreated specimens, are found in each reforming axopodium. Reforming axopodia without birefringent axonemes (and thus without microtubules) are never encountered. From these observations we conclude that the microtubules may be instrumental not only in the maintenance of the axopodia but also in their growth. Thus, if the microtubules are destroyed, the axopodia should retract and not reform until these tubular units are reassembled. During the cold treatment short segments of a 340-A tubule appeared; when the organisms were removed from the cold, these tubular segments disappeared. It seems probable that they are one of the disintegration products of the microtubules. A model is presented of our interpretation of how a 220-A microtubule transforms into a 340-A tubule and what this means in terms of the substructure of the untreated microtubules.