MEDICAL JOURNAL estimated chemically, it was of a low order, particularly in the intermediate ranges. One may conclude that single occasional smears are not accurate enough to give a useful indication of oestrogenic activity in this group of women.Even with sensitive chemical determinations, selection of those patients most likely to benefit from endocrine ablation is apparently not possible by means of OE, OD, and OT measurements alone. The trend shown by this small series suggests that the subjects with the larger amounts of the more active oestrogens, OE and OD, relative to the less active OT, may be the ones less likely to respond to castration. However, there is much overlapping, and the ratio OE plus OD :OT would not seem to have any practical predictive value.In most cases the operations were carried out as therapeutic procedures when metastases had actually occurred, rather than as a prophylactic measure. It might be that by the time these women were considered for gonadectomy, tumours which were possibly once hormone-dependent had become autonomous. Another possibility is that tumour growth is dependent not on the absolute amounts of oestrogens secreted but on the balance between the various oestrogens, their metabolites, and other hormones (Bulbrook, Greenwood, and Hayward, 1960 daily. This route of administration and the dosage were chosen as the method is the one used at this hospital. Nevertheless, the drug was poorly tolerated by many patients. Hirsuties, deepening of the voice, nausea, and vomiting often occurred, and sometimes hypertension and oedema developed in patients hitherto well controlled with adrenal substitution therapy. For these reasons testosterone treatment had often to be abandoned.Thiotepa was given by intramuscular injection, starting with 15 mg. three times a week for two weeks, and thereafter 15 mg. weekly. This scheme was chosen because most cases were treated as out-patients. Whitecell counts were initially performed twice weekly, and then weekly during maintenance therapy. Treatment was continued until leucopenia (defined as a total white-cell count of fewer than 4,000/c.mm.) rendered it hazardous or until deterioration was obvious. When patients developed leucopenia thiotepa was withheld until the white-cell count rose above 6,000/c.mm.; but usually there was a further fall in white cells, and eventually treatment with thictepa had to be stopped.Results Of the 12 patients studied, five had had a period of improvement after bilateral adrenalectomy and oophorectomy varying from 11 to 30 months. In the remaining seven progression of the malignant disease had continued in spite of operation. Thiotepa produced temporary improvement in two patients (Cases 5 and 7). In the remainder no effect was seen. Details of the individual cases are summarized in the Table, and the case histories of the patients who improved are given below.Case 5.-This patient underwent a radical mastectomy for carcinoma of the left breast in 1954, followed by a course of radiotherapy. She remained well unt...