1. An examination of eighty-nine samples of uncooked and thirty-eight samples of cooked food purchased by one particular hospital showed that the purchased food was bacteriologically clean.

2. A similar examination of 173 samples of food after it was cooked and prepared for serving in the hospital kitchens showed that, with the exception of cold chicken, the bacterial flora was greatly reduced by cooking. Ten of forty-six samples of cold chicken, however, were contaminated with fairly large numbers of Cl. welchii.

3. An investigation into the cooking and handling of the fowls indicated that contamination of the cooked fowls with kitchen dust was a probable explanation for the presence of Cl. welchii.

4. Immediate refrigeration of the fowls, well separated on shallow trays, was shown to be a satisfactory method of preventing the growth of contaminating Cl. welchii to any dangerous extent.

5. Six outbreaks of food poisoning, in which there was an association between cold chicken and the clinical symptoms of Cl. welchii food poisoning, are reported and discussed.

I have pleasure in thanking Dr Betty Hobbs of the Food Hygiene Laboratory of the Public Health Laboratory Service at Colindale for the serological typing of numerous strains of Cl. welchii; Mr D. B. Colquhoun for assistance with the phage-typing of the staphylococci; Mr G. Kerr for the photography; and the kitchen staff in the hospital concerned for making possible the numerous samplings of food and the other inquiries into the day-to-day affairs of the hospital catering department.