Abstract 1. (1) In continuation of our first paper ( Dastur et al. 1968) on the overall analysis of 1000 intracranial space-occupying lesions, this paper analyses in detail 370 gliomas which constituted 37.1% of the former but 48.2% of the 768 true neoplasms in this series. The average age for gliomas was 27 years, much younger than that of any reported series. 2. (2) Astrocytomas comprise 60.5% of gliomas. They have been classified into grades I, II, III and IV according to the degree of cellular anaplasia (Kernohan's method). Benign astrocytomas (grades I and II together) comprise 39.3% and the malignant (grades III and IV together) 60.7%. The average age for the former was 26 years, for the latter 35 years, and for the whole group 31.5 years. 3. (3) Of the total number of gliomas 27.3% (101) occurred in children under 15 years of age, a figure comparable with Japanese data but much higher than the figures reported from the West. Of these 41.5% were astrocytomas, of which nearly 70% were benign in contrast to 39% in the total material. Similarly, while two-thirds of all the total gliomas were supratentorial, two-thirds of the gliomas in childhood were infratentorial. Cerebellar astrocytoma was mainly responsible for these two features. 4. (4) Vascular and perivascular changes in astrocytomas were stressed and a case of combined astrocytoma and vascular sarcoma is described. The youngest age at onset (4 years) and longest survival (4 years) were observed in a boy with malignant astrocytoma. 5. (5) Ependymomas constituted the second most frequent type of glial tumour in both our total (14%) and paediatric (24%) material, being comparable to the incidence in Japan; but the proportion of our fourth ventricular ependymomas in children was higher than in Japan, and much higher than in any Western series. 6. (6) The medulloblastomas, almost equalling the proportion of ependymomas among children, were divided into the 3 histological subtypes of classical (49%), transitional (24%) and desmoplastic (27%) tumours. There was no significant age difference between the former, and the latter which was characterised by hemispheric location, surface extension and profuse reticulin in the fibroblastic part of the tumour. The histogenesis of the medulloblastoma is discussed. 7. (7) The optic nerve gliomas and neuro-hypophyseal gliomas were characterised by a common pattern of piloid astrocytoma. The former were all seen in children. The 4 latter constituted an unusually large proportion (1.1%) and occurred at all ages, and 1 of them showed excessive Rosenthal bodies. 8. (8) Four of the 7 pineal tumours appeared gliomatous with the pattern either of a pineocytoma or pineoblastoma. They are not very frequent tumours, thus contrasting with the findings in Japan. 9. (9) The choroid plexus papilloma also appears to be a tumour of childhood. One case with anaplastic change is described. 10. (10) Two cases of ganglioglioma were found. Both were seen in children, and were slow-growing, with astrocytic elements predominating. One case showed oligodendroglia and calcospherites. 11. (11) A most unusual medullo-epithelioma with a rich fibrovascular stroma is described. 12. (12) The effect of X-irradiation on a glioma is demonstrated through an unusual case of olfactory neuroblastoma which showed extensions into the frontal lobes and the naso-pharynx. An ependymoma operated on four times is also described to show the lack of cellular and nuclear outline which may result from irradiation. 13. (13) Cerebral reticulosis with its typical large mononuclear cells and cellular reticulin is demonstrated through the oldest case of our series, a man aged 76 years.