SummaryThe frequency of non-clostridial anaerobic infection was studied in 95 patients who had undergone acute appendicectomy: 49 received prophylactic metronidazole and 46 received placebo. Anaerobic infection did not develop in any of the metronidazole-treated patients, but infections did develop in nine (19" .) of the 46 controls. Metronidazole is conveniently administered by suppository to patients who cannot take oral drugs. Five patients with intra-abdominal infections caused by non-clostridial anaerobes were successfully treated with metronidazole.

Most workers agree that the cell that gives rise to colonies in the agar colony system is equivalent to the granulopoietic stem cell. 8 The formation of colonies, however, requires the activatton of this cell and the serial proliferation of its progeny, and we cannot distinguish from our data whether quinine had a toxic effect on either or both of these mechanisms.Our results encourage the further use of the marrow culture system for trying to identify in vitro drugs responsible tor agranulocytosis in particular patients. At present only watersoluble drugs can be tested, which will remain a problem until non-toxic solvents for insoluble drugs are developed. A further obvious but none the less important precaution is not to test parenteral preparations containing preservatives, which are well known to be inhibitory in the marrow culture system.

No abstract

The management of the cirrhotic patient with bleeding gastro‐oesophageal varices is discussed with particular reference to the role of emergency surgery. An analysis is made of the results obtained in 77 patients with emergency surgery to control recurrent haemorrhage. The deleterious effect of failing to control haemorrhage immediately and effectively in these cases is emphasized. If haemorrhage recurs in spite of conservative measures, including tamponade, surgery is considered mandatory. Only 8 cases have had an emergency portacaval anastomosis, with 4 survivors, all good‐risk cases. More patients might benefit from this operation but only if haemorrhage is rapidly and effectively brought under control in the first instance. The poor‐risk case has a 30 per cent chance of survival with a Boerema‐Crile operation. However, this is not a definitive operation and patients who survive this procedure should have a portacaval anastomosis as soon as their liver function has sufficiently recovered.

We report on two cases of spontaneous intramural rupture and haematoma formation in the oesophagus. This is an unusual and often misdiagnosed condition. Its diagnosis relies on the history and a barium swallow. Instrumentation can result in further damage to the oesophagus and should be avoided. Treatment is conservative and results in resolution of the haematoma and a return to normal swallowing.