SUMMARYOur light, and electron microscopic (EM) findings within the hypothalamic supraoptic (SO) and paraventricular (PV) nuclei of the normal female rabbit are in agreement with those reported earlier by other investigators for the same nuclei of the dog and rat. The neurons of these nuclei are the hypothalamic synthesis sites of the neurohypophyseal hormones.With the exception of the arcuate nucleus, none of the hypothalamic nuclei associated with the control of adenohypohpyseal function have been studied extensively with the electron microscope. On the basis of our EM findings within the female rabbit hypothalamus, all neurons observed within the preoptic (PO) and suprachiasmatic (SCH) nuclei of the non-mated control animal were morphologically identical to the conventional neuron as described by Peters, Palay and Webster (1970). However, following coitus, castration and laparotomy, many neurons of these nuclei showed subcellular changes that have been repeatedly associated with enhanced protein synthesis. These large ‘neurosecretory’ neurons were usually located near capillaries and characterized by their well developed Rough endoplasmic reticulum (RER) and Golgi profiles, dense populations of mitochondria and lysosomes and by the presence of a homogeneous population of densecore vesicles (DCV) showing a peak distribution of 120-140 nm. Since similar neurons were not observed within the PO and SCH of the normal control rabbit it is suggested that we were observing functional states of the same type of neuron and that these ultrastructural changes occur in response to endocrine manipulation.Two types of neurons described as ‘pale’ and ‘dark’ were observed within the arcuate nucleus of both the control and experimental female rabbit. Ultrastructurally, these neuron types were identical to those described by other investigators for the rat. It has been suggested that the ‘pale’ and ‘dark’ neurons of this hypothalamic nucleus represent functional states of the same type of cell. However, increases in the ratio of ‘dark’ to ‘pale’ neurons as observed within the arcuate nucleus of the rat following castration, were not seen in the rabbit. Similar findings were also not evident within the arcuate nucleus of the female rabbit following coitus.As far as could be determined, all neurons of the ventromedial (VMN) nuclei of both the control and experimental rabbit were morphologically identical to the smaller, conventional type neuron. Certainly, ultrastructural changes similar to those observed within the PO and SCH nuclei of the female rabbit following coitus, castration or laparotomy, were never observed.The basic zonation and subcellular organization of the female rabbit Median Eminence (ME) is similar to that described for other mammalian species. Our EM findings within the external layer of the rabbit ME, however, are not entirely in agreement with the earlier study of Duffy and Menefeef 1965). These investigators reported only one population of DCV within the axon terminals of the rabbit ME external layer. We feel that we have ultrastructural evidence for the presence of at least two distinct populations of DCV within this layer of the rabbit ME. Furthermore, since these vesicle populations occurred within separate axon profiles and terminals, differences in their content and origin are suggested.Certainly, the relationship between releasing factors (RF) and the various populations of DCV observed within the external layer of the mammalian ME is not well established. The smaller (90 nm - 100 nm) DCV we have observed probably contain the catecholamines, while those of larger (120 nm - 140 nm) diameters may well represent the carriers of the RF associated with gonadotropic activity. The latter view is based primarily on our finding or numerous ‘vesicle ghosts’ within the axon terminals abutting the perivascular space (PVS) of portal capillaries of rabbits sacrificed at 10 minutes post-coitus. The mean diameters of 137±14 nm obtained for these ghosts strongly supports the suggested depletion of only the larger of the two DCV populations. Similar changes were not apparent within the axon terminals containing homogenous populations of only the smaller DCV.Unquestionably, the precise hypothalamic synthesis sites for the RF associated with control of adenohypophyseal function, continues to provoke comment. From the results obtained from countless studies that have employed a variety of neuroendocrinilogical techniques, two main hypothalamic centers of RF synthesis have been suggested: a) the medial basal hypothalamus (MBH) or hypophysiotropic area (HTA) and b) the anterior hypothalamus. The ultrastructural studies carried out to date within this laboratoiy are in favour of the latter for the following reasons:1) — the presence of large DCV and ‘vesicle ghosts’ within the external layer of the rabbit ME with diameters similar to those of the large (120-150 nm) DCV synthesized within the PO and SCH nuclei of the same animal in response to coitus, castration and laparotomy.2) — the absence of evidence for the storage of these large DCV within the somata of PO and SCH nuclei, suggesting their immediate transport toward the ME.3) — the absence of any ultrastructural changes within neuron somata of the rabbit arcuate nuclei which might reflect enhanced neurosecretory activity in response to coitus and/or castration.These ultrastructural findings within the rabbit hypothalamus may, therefore, provide the first evidence of a morphological nature for the actual release of RF from their ME storage sites, as well as their synthesis within certain neurons of the anterior hypothalamus.