THE obvious structural and functional interdependence of normal tissue stroma and malignant cells in solid tumours persuaded the earliest radiobiological investigators that the effect of ionising radiation on these composite structures was the resultant of separate damage to the normal and malignant tissue components. It is understandable that, with a persistent inability to quantitate separately the damage to either component, unlimited scope prevailed for the assertion of rival theories in which the response of a tumour was attributed preferentially to direct damage to one or other component. Histological study of irradiated experimental tumours is of very limited value in assessing the contribution of direct stromal damage. This follows from the fact that direct damage to blood vessels is not readily distinguishable from changes consequent on the regression of stroma which must be expected to follow the dissolution of tumour cells whose reproductive integrity has been directly damaged by the irradiation.Recent developments in the quantitative radiobiology of mammalian tumour cells irradiated in vivo have encouraged interpretations of tumour response which refer, often exclusively, to the direct effect of radiation on the clonogenic cell population of the tumour. Such exclusive consideration has undoubtedly been proved to be justified in respect of the relation between the estimated size of a tumour cell population in a tumour and the single dose of radiation required for its cure under specified conditions of oxygenation. This relation has been found to accord with the predictions of relevant radiation survival curves (Hewitt, 1963;Reinhold and De Bree, 1966), this last information being obtained under conditions where stromal changes make no contribution. Suit, Shalek and Wette (1964) conclude from their extensive dose-cure studies of murine adenocarcinoma that their results " do not indicate a tissue effect on cellular radiosensitivity, tumour bed effect on tumour curability, or non-specific host-tumour effect ".If, as it appears, radiation-induced damage to the stroma makes no measurable contribution to the eradication or " cure " of a tumour, the question remains whether such damage influences the character of the changes of tumour volume which are brought about by irradiation of a tumour in vivo. The consideration achieves particular importance when attempts are made to interpret tumour regrowth curves in terms of survival curves for the clonogenic tumour cells. Thomlinson and Craddock (1967) state that, for the rat fibrosarcoma they studied, the oxygen enhancement ratio determined from measurements of the growth response of their tumours to irradiation is considerably greater than that obtained from in vitro studies of the clonogenic cells of their tumour. They refer to capillary damage as possibly conducing to the discrepancy. Several reports have appeared of the volume response of tumours to a single or fractionated dose of