classified by them as 'new and independent carcinomas '. We admit that if such a ' transition zone ' were present this would be evidence in support of the tumour being a 'new and independent carcinoma ', but in many obvious primary carcinomas no ' transition zone ' can be detected. Therefore, although a closer study of the histology of the edge of a tumour may provide valuable information in some cases, it is unsafe to rely on this alone when attempting to distinguish between primary and secondary tumours.