The friction coefficients measured in diarthrodial joints are small. Theories of joint lubrication attribute this efficiency to entrapment or movement of synovial fluid, yet anatomical models of the surface are based on studies of isolated fragments of cartilage, not functional joints. To investigate the functional interrelationship of joint surfaces and synovial fluid, the ultrastructure of loaded joints was examined. Twenty-four New Zealand white rabbit knee joints were loaded either statically or moved ex vivo using simulated muscle forces and then plunge-frozen under load. After fixation in the frozen/loaded state by freeze-substitution fixation, the medial joint compartments were embedded in epoxy resin while still articulated. Bone was trimmed away from the articular surfaces, permitting the cartilage to be sectioned for light and electron microscopy. These joint surfaces were then compared with controls which were not loaded, not moved or had been disarticulated prior to embedding. Articular surfaces of loaded joints were smooth at magnifications from ×35 to ×7500, whereas the tibial surfaces of nonloaded joints were irregular. Small pools of joint fluid were observed at the meniscal edge and beneath the anterior horn of the meniscus. At magnifications of ×40000, the joint surfaces were separated by a uniform 100 nm space containing fluid. An amorphous, electron dense articular surface lamina was present but, when loaded, was thicker and flatter than previously reported. No surface pits or bumps were visible in embedded, loaded joints. This is the first ultrastructural study of intact loaded joints. The findings suggest that fluid film lubrication is present in diarthrodial joints, but the fluid sequestration postulated in several models is not apparent.

The aim of this study was to examine the kinetics of neovascularisation of splenic autoimplants into the abdominal cavity after splenectomy in mice. Sixty-eight female Swiss mice were submitted to splenectomy. The spleen from each animal was sliced and the slices were implanted into the abdominal cavity. Groups of animals were killed after 1, 2, 3, 5, 7, 15, 21, 28, 42, 56, 70 and 84 d. Fluorescent polystyrene microspheres were injected via the orbital venous plexus before killing and the splenules were removed 5 min later for light and electron microscopy. Mesenteric blood vessels were injected with coloured latex to study the origin of the nutrient vessels. Three days after the implant the microspheres were observed at the periphery and then migrating to the internal parts of the implant in the subsequent days. The blood supply to the implants originated from branches of the splenic, short gastric, mesenteric and gastroepiploic arteries. It is concluded that revascularisation of splenic autografts proceeds centripetally, starting as early as 3 d after implantation.

Eight normal human spinal cords were studied for motoneuron (Mn) groups and columns. Spinal segments (C1 to Coc.) were identified and embedded in paraffin wax. Serial cross sections were cut at 25 μm and stained by cresyl violet. Cross-sectional profiles of the spinal cord were traced for each segmental level and the outlines of the various Mn groups superimposed. These charts (maps) were used to examine intra and intersegmental changes in the relative positions of the columns. An attempt was made to provide topographical picture of Mn groups of individual segments. In the cervical region neuronal groups were more numerous but smaller and less distinct, while in the lumbosacral region they were fewer, larger and at many levels better circumscribed. The average number of Mn groups at any segmental level was 3–4 and never exceeded 5. C4, C5, C6, C7, L4, L5 and S1 contained numerous Mn groups. Maximum intrasegmental changes were noted at C3, C4, C7, T1, and S2, while at C5, C6, all thoracic, L1 L2 and L3, the pattern was constant throughout the segment. Eleven motor columns were traced in the human spinal cord. Column 1 belonged to the medial division and columns 2–11 to the lateral division of the ventral grey horn. Columns 1 and 2 were the most extensive as they were traceable from the lower medulla to S3 segment. Columns 3–8 were confined to cervical segments (including T1), while columns 9–11 were traced in lumbosacral segments. In general, motor columns followed a definite mode for their appearance and disappearance. Many of them showed rotation from a dorsal to a ventromedial direction.

The purpose of the present study was to analyse the replicability of the scoring of discontinuous traits. This was assessed on a sample of 100 skulls from the Frassetto collection (Dipartimento di Biologia Evoluzionistica Sperimentale of Bologna University) analysed through intraobserver comparisons: the discontinuous traits were determined on the same skulls and by the same observer on 3 separate occasions. The scoring was also assessed through interobserver comparisons: 3 different observers performed an independent survey on the same skulls. The results show that there were no significant differences in the discontinuous trait frequencies between the 3 different scorings by the same observer, but there were sometimes significant differences between different observers. Caution should thus be taken in applying the frequencies of these traits to population research. After an indispensable control of material conditions (subject age included), consideration must be given to standardisation procedures between observers, otherwise this may be an additional source of variability in cranial discontinuous trait scoring.

The results obtained using morphometric variables which describe fin ray regeneration patterns are reported for individual fin ray amputations in the goldfish (Carassius auratus) and zebrafish (Brachydanio rerio). Classical and updated experiments are compared to verify previous morphogenetic models of cell tractions (Oster et al. 1983) or epidermis-mesenchyme induction (Saunders et al. 1959) applied to the limb of other vertebrates. Position-dependent patterns within the fin of Carassius auratus are analysed under a comparative protocol using morphometric methods. Conditions in which the apical epidermis is separated from blastema may differentiate small fin rays, thus suggesting this epidermis is involved in blastemal formation. Blastemal cells differentiating as lepidotrichia forming cells (LFCs) may also be related to morphological changes in covering epidermis. Long-range interactions from neighbouring fin ray blastemas or short-range interactions within the blastema, may be postulated through the analysis of segmentation.

In the epiphysis of long bones, different patterns of development of ossification processes have been described in different species. The development of the vascularisation of the femoral head has not yet been fully clarified, although its role in the ossification process is obvious. Our aim was to investigate ossification and vascular proliferation and their relationship, in growing rat femoral heads. Male Wistar rats aged ∼ 1, 5 and 8 wk and 4, 8 and 12 mo were used. Light microscopy frontal sections and vascular corrosion casts observed by scanning electron microscopy were employed. In the rat proximal femoral epiphysis, ossification develops from the medullary circulation of the diaphysis, quickly extending to the neck and the base of the head. Hypertrophic chondrocytes occupy the epiphyseal cartilage, and a physeal plate with regular cell columns is present. Starting from about the end of the third month one or more points of fibrovascular outgrowth, above the physeal line, can be observed in each sample. They are often placed centrally or, sometimes, peripherally. The fibrovascular outgrowths penetrate deeply into the cartilage and extend laterally. At age 8 mo, large fibro-osseous peduncles connect the epiphysis to the diaphyseal tissue. At 12 mo, the entire epiphysis appears calcified with an almost total absence of residual cartilage islands. This situation differs in man and in other mammals due both to differing thickness of the cartilage and to the presence of more extensive sources of blood vessels other than the diaphyseal microcirculation, as supplied by the teres ligament and Hunter's circle. In young rats, subchondral vessels and the synovial fluid could play a role in feeding the ossifying cartilage. Later, a loss of resistance of the physis due to marked degeneration of the cell columns, and extensive chondrocyte hypertrophy permit fibrovascular penetration starting from diaphyseal vessels rather than neighbouring vascular territories, such as those of the periosteum and capsule.

This study confirms that the median artery may persist in adult life in 2 different patterns, palmar and antebrachial, based on their vascular territory. The palmar type, which represents the embryonic pattern, is large, long and reaches the palm. The antebrachial type, which represents a partial regression of the embryonic artery is slender, short, and terminates before reaching the wrist. These 2 arterial patterns appear with a different incidence. The palmar pattern was studied in the whole sample (120 cadavers) and had an incidence of 20%, being more frequent in females than in males (1.3[ratio ]1), occurring unilaterally more often than bilaterally (4[ratio ]1) and slightly more frequently on the right than on the left (1.1[ratio ]1). The antebrachial pattern was studied in only 79 cadavers and had an incidence of 76%, being more frequent in females than in males (1.6[ratio ]1); it was commoner unilaterally than bilaterally (1.5[ratio ]1) and was again slightly more prevalent on the right than on the left (1.2[ratio ]1). The origin of the median artery was variable in both patterns. The palmar type most frequently arose from the caudal angle between the ulnar artery and its common interosseous trunk (59%). The antebrachial pattern most frequently originated from the anterior interosseous artery (55%). Other origins, for both patterns, were from the ulnar artery or from the common interosseous trunk. The median artery in the antebrachial pattern terminated in the upper third (74%) or in the distal third of the forearm (26%). However, the palmar pattern ended as the 1st, 2nd or 1st and 2nd common digital arteries (65%) or joined the superficial palmar arch (35%). The median artery passed either anterior (67%) or posterior (25%) to the anterior interosseous nerve. It pierced the median nerve in the upper third of the forearm in 41% of cases with the palmar pattern and in none of the antebrachial cases. In 1 case the artery pierced both the anterior interosseous and median nerves.

The aims of the present study were: (1) to supply further knowledge about variations in nonmetric cranial traits in relation to sex, age and laterality and (2) to evaluate biological distance between samples from a recent population. The incidence of 18 nonmetric variants of the cranium were determined in 3 adult samples of 394 skulls of known sex from North Sardinia (Sassari, Alghero and Ozieri); for the Sassari sample (n = 200) age at death was also known. Some significant sex differences were observed. Age did not appear to influence the frequency of the discontinuous traits but did for legibility. Side differences may provide important information about environmental influences. The interpopulation analysis indicates a stronger relationship between samples that are geographically closer (Sassari and Alghero), in accordance with other studies, strengthening the hypothesis of the validity of the use of nonmetric traits in the study of the peopling of a territory.

Information abut the existence and anatomy of arterial anastomoses with the porcine bronchial artery is lacking in the literature. Prior to basic physiological investigations in a porcine model related to lung transplantation with bronchial artery revascularisation, this study was designed to examine the anatomy of systemic arterial anastomoses with the bronchial artery system. Twenty pigs were studied in 3 groups. In 2 groups the heart–lung block was removed with all mediastinal structures. One group served for investigation of coronary–bronchial artery anastomoses and one for investigation of oesophageal–bronchial artery anastomoses. The systemic arteries to be examined were cannulated. The inflated heart–lung block was examined macroscopically with Evans blue, and radiographically after contrast injection. In the 3rd group intercostobronchial artery anastomoses were studied radiographically with the heart–lung block in situ. Coronary–bronchial artery anastomoses were demonstrated in 3 of the 5 pigs with an aortic ‘pouch’ technique, but contrast was very limited in 2 of these 3. Oesophageal arterial anastomoses with bronchial arterial branches and/or the pulmonary veins were demonstrated in 6 of the 7 pigs and more markedly than the coronary–bronchial anastomoses. Intercostobronchial artery anastomoses could not be demonstrated angiographically. It was concluded that the existence of coronary–bronchial and oesophageal–bronchial artery anastomoses in the pig appear to establish an arterial net between the base of the heart and the distal oesophagus. The resemblance to human oesophageal–bronchial artery anastomoses supports use of a porcine model for experimental studies.

Qualitative and quantitative information is reported on the morphological changes that occur in nerve fibres and nonneuronal cells of peripheral nerve during the lifetime of the mouse. Tibial nerves of mice aged 6–33 mo were studied. With ageing, collagen accumulates in the perineurium and lipid droplets in the perineurial cells. Macrophages and mast cells increase in number, and onion bulbs and collagen pockets are frequently present. Schwann cells associated with myelinated fibres (MF) slightly decrease in number in parallel with an increase of the internodal length from 6 to 12 mo, but increase in older nerves when demyelination and remyelination are common. The unmyelinated axon to myelinated fibre (UA/MF) ratio was about 2 until 12 mo, decreasing to 1.6 by 27 mo. In older mice, the loss of nerve fibres involves UA (50% loss of 27–33 mo cf. 6 mo) more markedly than MF (35%). In aged nerves wide incisures and infolded or outfolded myelin loops are frequent, resulting in an increased irregularity in the morphology of fibres along the internodes. In the mouse there is an adult time period, 12–20 mo, during which several features of degeneration progressively appear, and an ageing period from 20 mo upwards when the nerve suffers a general disorganisation and marked fibre loss.

The gross and microscopic anatomy of the Cape fur seal heart, lung, liver, spleen, stomach, intestine and kidneys (n = 31 seals) is described. Absolute and relative size of organs from 30 male seals are presented, with histological examination conducted on 7 animals. The relationship between log body weight, log organ weight and age was investigated using linear regression. Twenty five animals were of known age, while 6 were aged from counts of incremental lines observed in the dentine of tooth sections. For the range of ages represented in this study, body weight changes were accurately described by the exponential growth equation, weight = wort, with body weight increasing by 23% per annum until at least 9–10 y of age. Organ weight increased at a rate of between 25% and 33% per annum until at least 9–10 y of age, with the exception of the intestines, where exponential increase appeared to have ceased by about 7 y. The relationship between body weight and organ weight was investigated using logarithmic transformations of the allometric equation, y = axb, where the exponent b is 1 if organ weight is proportional to body weight. Most organs increased in proportion to the body. However, the heart, liver and spleen had exponents b > 1, suggesting that these organs increased at a faster rate than the body. The basic anatomical features of the viscera were similar to those of other pinnipeds, with some exceptions, including the arrangement of the multilobed lung and liver. Apart from the large liver and kidneys, relative size of the organs did not differ greatly from similar sized terrestrial carnivores. The histological features of the organs were generally consistent with those previously described for this species and other otariids. The heart, as in other pinnipeds, was unlike that of cetacea in not having unusually thick endocardium or prominent Purkinje cells. Notable histological features of the lungs included prominent fibrous septa, prominent smooth muscle bundles, cartilage extending to the level of the alveolar sacs and ample lymphoid tissue. The spleen had a thick capsule, well developed trabeculae and plentiful plasma cells. Abundant parietal cells were present in the fundic glands and lymphoid follicles were present in the gastric lamina propria, particularly in the pyloric region. Small intestinal villi were very long but this could have resulted from underlying chronic inflammation. Lymphoid follicles were prominent in the colon. The kidney reniculi each had a complete cortex, medulla and calyx, but a sportaperi medullaris musculosa was not identified.

Vein-to-artery graft surgery is used widely to by-pass arterial stenoses, but such grafts can fail over a prolonged period as a result of excessive neointimal hyperplasia causing thrombosis and graft occlusion. It has been suggested that neointimal hyperplasia, in vein grafts, is a result of the vessel wall adapting to the higher intraluminal pressure of the arterial circulation, compared with the venous circulation. Autologous artery grafts have been used to bypass arterial stenoses. Initially it was assumed that donor artery segments would not develop neointimal hyperplasia as they are already adapted to the arterial circulation but this is not so. In this study we postulated that surgical or postsurgical trauma was the cause of neointimal hyperplasia in autologous artery-to-artery grafts. In addition, as artery grafts are pre-adapted to the arterial circulation, autologous artery-to-artery grafts in hypertensive rats should develop similar levels of neointimal hyperplasia as seen in normotensive rats. Artery-to-artery grafts were placed in a series of 20 spontaneously hypertensive rats (SHR). In a separate series of sham grafting experiments the effects of anoxia and clamp trauma were assessed in SHR and WKy normotensive control rats. Finally, clamping, anoxia and anastomosis trauma were assessed in a similar series of rats. In the artery-to-artery graft series there was no difference in neointimal thickness between the SHR and that previously reported for normotensive rats. Minimal neointimal hyperplasia was demonstrated in the sham grafted series of rats and only slightly more in the single anastomosis series. It was only in the full grafting procedure that considerable neointimal hyperplasia developed. These data demonstrate that neointimal hyperplasia in artery-to-artery grafts is not exacerbated by the hypertension. In addition, trauma appears to be the initiator of neointimal hyperplasia and the extent of trauma correlates with the degree of neointimal hyperplasia.

The study was designed to reveal the distribution of various mast cell types in the forebrain of the adult sheep, hedgehog and rat. Based on their histochemical and immunocytochemical characteristics, mast cells were categorised as (1) connective tissue-type mast cells, staining metachromatically purple with the toluidine blue method, or pale red with the Alcian blue/safranin method, (2) mucosal-type or immature mast cells staining blue with the Alcian blue/safranin method and (3) serotonin immunopositive mast cells. All 3 types of brain mast cells in all species studied were located in both white and grey matter, often associated with intraparenchymal blood vessels. Their distribution pattern exhibited interspecies differences, while their number varied considerably not only between species but also between individuals of each species. A distributional left-right asymmetry, with more cells present on the left side, was observed in all species studied but it was most prominent in the sheep brain. In the sheep, mast cells were abundantly distributed in forebrain areas, while in the hedgehog and the rat forebrain, mast cells were less widely distributed and were relatively or substantially fewer in number respectively. A limited number of brain mast cells, in all 3 species, but primarily in the rat, were found to react both immunocytochemically to 5-HT antibody and histochemically with Alcian blue/safranin staining.

The human vomeronasal organ is of interest because of its potential role in sex pheromone detection. Due to the scarcity of early human material, studies of its development have concentrated on fetal rather than embryonic stages. The availability of embryonic specimens in the Walmsley Collection has enabled us to study the development of the vomeronasal organ (VNO) in human embryos between Carnegie Stages 17 and 23. Embryos at Carnegie Stage 17 or below showed no evidence of a VNO. One embryo with characteristics intermediate between Carnegie stages 17 and 18 was the earliest to show evidence of a VNO, in the form of a shallow indentation. All embryos at Carnegie Stages 18 or later had VNOs. Three-dimensional computer reconstructions were made of the VNO in each specimen where this was possible. This in part depended on the plane of section. The total volume and lumen volume were measured from these reconstructions and the volume of the vomeronasal epithelium was calculated by subtraction. A generally consistent increase in total volume and epithelial volume was observed with increasing developmental stage. The lumen contributed rather little to the total volume at these stages.

Sympathetic nerve-mediated vasoconstriction in iris arterioles of mature rats occurs via the activation of α1B-adrenoceptors alone, while in immature rat iris arterioles, vasoconstriction occurs via activation of both α1- and α2-adrenoceptors. In mature rats the vast majority of sympathetic varicosities form close neuroeffector junctions. Serial section electron microscopy of 14 d iris arterioles has been used to determine whether restriction in physiological receptor types with age may result from the establishment of these close neuroeffector junctions. Ninety varicosities which lay within 4 μm of arteriolar smooth muscle were followed for their entire length. Varicosities rarely contained dense cored vesicles even after treatment with 5-hydroxydopamine. 47% of varicosities formed close associations with muscle cells and 88% formed close associations with muscle cells or melanocytes. Varicosities in bundles were as likely as single varicosities to form close associations with vascular smooth muscle cells, although the distribution of synaptic vesicles in single varicosities did not show the asymmetric accumulation towards the smooth muscle cells seen in the varicosities in bundles which were frequently clustered together. We conclude that restriction of physiological receptor types during development does not appear to correlate with the establishment of close neuroeffector junctions, although changes in presynaptic structures may contribute to the refinement of postsynaptic responses.

Angiotensin II (ANG II) is increasingly recognised as a growth factor, both in its own right and through interactions with other growth factors. There is a high density of ANG II receptors in the rat fetus, especially the AT2 receptor, the function of which is still uncertain. We have now studied the effects of ANG II on growth and development in the rat embryo in vitro between d 9.5 and 11.5, and characterised the receptor subtype mediating these effects. Embryos were cultured in whole rat serum, a high molecular weight retenate after ultrafiltration of whole rat serum, retenate with angiotensin II and retenate with ANG II and AT1 or AT2 receptor blockers. Growth and development were scored using conventional methods. Culture in retenate was associated with a marked reduction in growth and development by comparison with whole rat serum. This was partly, and significantly (P<0.001), reversed by angiotensin II. The optimum concentration of angiotensin II was found to be angiotensin II 10−11M, within the physiological range. Angiotensin II had highly significant effects on both somatic (P<0.001) and yolk sac/allantoic (P<0.005) development. The latter effects suggest a role for angiotensin II in placentation. The effects of angiotensin II were blocked by PD123319, an AT2 blocker, but not by GR117289, an AT1 blocker. Interestingly, culture in retenate with GR117289 without added angiotensin II was also associated with some increase in growth (P<0.05). Angiotensin II in low concentrations was measurable in the retenate, presumably arising from the action of endogenous renin on angiotensinogen. We therefore postulate that this effect of GR117289 was due to the action of endogenous angiotensin II on ‘uncovered’ AT2 receptors. This study has thus demonstrated a direct growth promoting effect of angiotensin II during organogenesis in the whole rat embryo in vitro. This effect is mediated through the AT2 receptors.

Adult dorsal root ganglion neurons express oligosaccharides conjugated to lipids that may be involved in cell–cell recognition, and consequently in the laminar organisation of their central terminations. This paper describes an immunohistochemical study of the developmental expression of 2 lactoseries (LA4 and LD2) and 1 globoseries (SSEA4) oligosaccharide conjugates in rats from embryonic d 19 to postnatal d 60. The expression of calcitonin gene related peptide and the growth associated protein GAP43 was also examined for comparative purposes. We found that these oligosaccharide conjugates begin to be expressed after birth, suggesting that they may be involved in maturation of the central or peripheral terminations, rather than axonal guidance.

The vomeronasal organ (VNO) is a chemoreceptive structure that has not been extensively studied in the Xenarthran order. Tissue samples from the VNO of the armadillo Chaetophractus villosus were prepared for light and electron microscopy. The VNO is located in the anterior part of the base of the nasal septum. It is tubular in shape, ∼ 18 mm in length and opens in the rostral region of the nasal cavity and with a blind caudal end. Its lumen is lined by sensory (SE) and nonsensory (NSE) epithelium. The SE shows sensory, supporting and basal cells whereas the NSE contains ciliated and nonciliated secretory cells and basal cells. At the ultrastructural level, the sensory cells appear as bipolar neurons with conspicuous microvilli on their free surface. The supporting cells of the SE contain numerous membrane-bound vesicles in their apical regions. A peculiar feature not found in other mammals, is the presence of concentric whorls of RER cisterns frequently observed in their basal expansions. Infiltrating plasma cells can be detected in the SE basal region close to the dorsal junctional area. This region also exhibits an unusual type of basal cell, probably responsible for the generation of new vomeronasal receptor neurons. The ciliated NSE cells exhibit numerous ovoids or irregularly shaped membranous protrusions projecting from the plasma membrane of the cilia. As far as we know, this is the first study reporting the presence of this feature in ciliated NSE cells. The nonciliated cells are characterised by scarce large secretory granules and apical microvilli. The vomeronasal glands are compound-branched tubuloacinar glands with serous acinar cells. Four types of secretory granules are present. The ducts of these glands reach the lumen in the dorsolateral region between the NSE and SE. Hypolemmal nerve terminals were observed contacting secretory cells. Fenestrated and nonfenestrated capillaries constitute the vascular supply to these glands. Plasma cells, intimately associated with acinar cells, were frequently observed.

The streptozotocin (STZ)-diabetic rat, the most commonly employed model of experimental diabetic neuropathy, is characterised by a reduction in nerve conduction velocity, pain threshold and blood flow. Whether or not structural abnormalities underlie these functional abnormalities is unclear. 10 adult male Sprague–Dawley STZ-diabetic rats (diabetes duration 27 d) and 10 age-matched (23 wk) control animals were studied. Motor nerve conduction velocity (m s−1) was significantly reduced in diabetic (41.31±0.8) compared with control (46.15±1.5) animals (P<0.001). The concentration of sciatic nerve glucose (P<0.001), fructose (P<0.001) and sorbitol (P<0.001) was elevated, and myoinositol (P<0.001) was reduced in diabetic compared with control animals. Detailed morphometric studies demonstrated no significant difference in fascicular area, myelinated fibre density, fibre and axon areas as well as unmyelinated fibre density and diameter. Endoneurial capillary density, basement membrane area and endothelial cell profile number did not differ between diabetic and control animals. However, luminal area (P<0.03) was increased and endothelial cell area (P<0.08) was decreased in the diabetic rats. We conclude there is no detectable structural basis for the reduction in nerve conduction velocity, pain threshold or blood flow, observed in the streptozotocin diabetic rat.

Neuronal migration is guided by long radially oriented glial fibres. During late stages of development radial glial cells are transformed into astrocytes. A predominant intermediate filament protein within radial glial cells and immature astrocytes is vimentin. In this study fetal brain sections were used to demonstrate the transient features of vimentin-positive radial glia. In the lower half of the cerebral wall of the 6th gestational month bundles, curvature, and crossing of vimentin-positive fibres are regularly seen. Moreover, fibres terminating on vessels are observed. In the upper half fibres are radially oriented; when ascending towards the pial surface the number and diameter of fibres appears conspicuously decreased. Radially aligned fibres display numerous varicosities. In the 8th month the bulk of vimentin-positive fibres is encountered next to the ganglionic eminence and below isocortical cerebral fissures. The dentate gyrus is conspicuous due to its high amount of immunolabelled fibres. Furthermore, densely packed fibres are visible within the internal and external capsule and in the vicinity of the anterior commissure. Radial glial somata are found in the proliferative areas as well as in the adjacent white matter. In the latter location bipolar, monopolar and stellate vimentin-positive cells are present. The results demonstrate an area-specific distribution pattern of vimentin-positive structures which can be correlated with migrational events. Areas maturing late in development for instance, reveal dense immunolabelling in the 8th month. The orientation and position of radial fibres point to an additional developmental role of these fibres, i.e. their involvement in the guidance of growing axons. Moreover, the arrangement and morphology of vimentin-positive fibres, such as retraction of fibres or occurrence of varicosities, are indicative of degenerative events. Accordingly, a transformation of radial glial somata, their displacement towards the white matter and finally the growth of stellate processes can clearly be demonstrated.