THE treatment of intracranial tumours by any method a t present conceived presents certain problems peculiar t o this field of surgery. There are difliculties of diagnosis, both anatomical and pathological ; difficulties of access ; and complications which arise from sudden changes in tension within the dural capsule, with resultant alterations in the circulation and damage t o the cerebral tissue. Gradually these problems havc bcen defined and tackled, and modifications in technique have been introduced to improve the methods of approach t o the brain, to obviate dangerous changes in intracranial tension, to reduce the risk of infection to the minimum, and to control haemorrhage. The recent adaptation of electrosurgical methods, the exact value of which is a t present difficult t o estimate, has been acclaimed the most notable advance in surgery since the introduction of antiseptics.Technical advances have led t o fresh discoveries in regard to the pathology of brain turnours, and have rendered possible and successful procedures which before the beginning of the present century would have proved too hazardous. This applies especially to the treatment of encapsulated tumours, but in the case of the gliomas it must be confessed that attempted removal, although carrying a much lower immediate mortality-rate than formerly, gives but a small chance of ultimate cure. Seeing that, if pituitary tumours be excluded, the glioma group constitutes over 70 per cent of the primary brain tumours verified by operation, the outlook is rather discouraging, and the hope has been expressed that radiotherapy might succeed where surgery has failed. The present studies have been undertaken t o try t o discover how far this hope is justifiable. The investigation is confined to the glioma group, though there are many recorded observations in regard to irradiation of endotheliomas, pituitary tumours, and neuroiibromas.
PATHOLOGICAL CONSIDERATIONS.In order to understand the problem of the gliomas it is necessary t o make a brief reference t o their pathology. Of the many criteria which indicate the malignancy of a tumour, the most important is the character of the neoplastic cells, and especially the degree to which they are differentiated. The more embryonic and undifferentiated the cells, the more active is the growth and the more malignant the tumour ; and there is evidence that the