1. A bacteriological laboratory was provided in the medical departments of two passenger ships—a modern luxury liner cruising in the Mediterranean for 1 month and an older vessel, now out of commission, on a 3 months round voyage through the tropics. Galley hygiene in relation to gastro-enteritis was investigated on both ships.

2. 125 samples of water and iced water were examined. Samples of water chlorinated on the ships were usually satisfactory. Many port waters sampled from the barge or hosepipe were contaminated—some mildly, some profusely. A fault in an ice-making machine led to pollution of ice used for various purposes. Swimming-bath water was usually mildly polluted only; occasionally in the tropics when the baths were very popular the count rose to 1600 Esch. coli per 100 ml.

3. Thirty-five samples of milk, ice-cream and churn washings were examined. English liquid pasteurized milk stored in the cold gave satisfactory results, but the rehydrated dried milk used exclusively on the round voyage through the tropics and swabs from the apparatus used in its preparation were contaminated with Esch. coli. Ice-cream samples from ship I gave satisfactory results, and of two samples of ice-cream powder examined on ship II one was satisfactory and the other gave a poor result. Occasional churns which had been superficially cleaned but not sterilized had a high general and Esch. coli count.

4. 189 samples of food were examined, twenty-two for general and coliform counts only. Results were variable according to foodstuff and atmospheric temperature. Some cold cooked meatstuffs gave low counts and absence of coliform bacilli in 1/10 g.; others had plate counts of 300,000 to 25 million per gram with Esch. coli in 1/1000 dilution, e.g. crayfish ready for the table. Salmonellae were not found in any samples but occasionally small numbers of coagulase-positive staphylococci and non-haemolytic Cl. welchii were isolated.

5. Sixteen samples of imitation cream and washings from savoy bags were examined. Samples from freshly opened cans of cream gave satisfactory results. In whipped cream and in cream from cakes and sweets, Esch. coli was found in 1/1000 dilutions, and rinsings from savoy bags in use gave probable numbers of Esch. coli of 18,000 + / 100 ml. Counts were still high after the bags had been washed and dried and they were far from sterile even after ‘boiling’ and drying.

6. Twenty-three samples of wash water from salad vegetables and fruit had high counts of Esch. coli. After these articles had been washed in water containing 80 p.p.m. sodium hypochiorite, coliform organisms were not found in 100 ml. of water. Potassium permanganate was considered to be of doubtful value. The results of laboratory experiments to confirm the concentration of hypochiorite and time necessary to destroy Esch. coli on lettuce and watercress are shown in Tables 3 and 4.

7. Results from forty-four samples of wash and rinse waters from dish-washing machines and one bowl showed that the temperatures were often too low; with temperatures above 50° C. coliform bacilli were absent in 1 ml. quantities, but waters at temperatures below 42° C. gave probable Esch. coli counts of 1800 + / 100 ml. The bacteriological condition of the crockery and cutlery varied according to the wash and rinse waters. Rinses from metal dish covers gave poor results, with 1800+ Esch. coli per 100 ml. of rinse and occasionally Cl. welchii and coagulase-positive staphylococci. Wash and rinse waters, with temperatures of 35°ndash;42° C. used for pots and pans gave probable Esch. coli counts of 1800 + /100 ml.

8. Rinse waters from twenty cloths used in the galleys, pantries and dining rooms were nearly all heavily contaminated with Esch. coli, with probable numbers of 18,000 + /100 ml.; these counts were reduced to negligible figures by soaking the cloths in chloride of lime.

9. Swabs were taken from thirty-eight cutting-boards and twenty-six surfaces of sinks, shelves and miscellaneous articles. Many boards were heavily contaminated, the probable number of Esch. coli being 18,000 + /100 ml. of diluent; lecithinase-positive clostridia were found also. The boards could be sterilized by brushing with a stout wire brush followed by chlorination.

Other surfaces, and swabs from various knives in the kitchen, including the blade of the slicing machine, frequently gave high counts of Esch. coli.

10. 329 swabs from nostrils and fingers of saloon and galley staffs yielded one cutlure of coagulase-positive staphylococci from 111 nasal swabs and three cultures of coagulase-positive staphylococci and seven of Esch. coli from 218 finger swabs from twenty-seven men.

11. Twenty-eight faecal specimens were examined from fifty-four cases of gastro-enteritis reported on the Mediterranean cruises; salmonellae were not found, but heat-resistant Cl. welchii was isolated from six samples (21%), and the clinical picture was that of Cl. welchii food poisoning.

12. 121 faeces samples from 137 patients with gastro-enteritis were examined chiefly for salmonellae and coagulase-positive staphylococci on the outward leg of the round voyage through the tropics. Salm. typhimurium was isolated from a clinical case and Cl. welchii from one typical case of food poisoning. On the return journey all samples were examined also for Cl. welchii; of thirty-five stool samples from thirty-five patients, six (17%) yielded lecithinase-positive clostridia; salmonellae were not found.

13. Heat-resistant strains of Cl. welchii were also isolated from two chopping-boards, a knife, cold beef, brawn, two wiping cloths, a dish cover, two samples of pie and corned beef.