The effectiveness of a representative range of fabrics in restricting dispersal through them of dry skin-borne bacteria has been examined. The fabrics were tested made up into trousers which were worn by volunteers during standardized exercise in a test chamber operated within a unidirectional flow clean-air room. Under these conditions, with careful attention to sealing at ankles and waist, it was possible to estimate penetration as low as 0·3%. Penetrations as low as 1% were observed with some synthetic fabrics. These had a relatively high surface resistivity and developed significant electrostatic charges.

When the observed values for penetration were compared with the results of a series of measurements and tests made on the fabrics it was clear that the correlation between these values and the other results was in every case very close for all the five woven cotton or cotton terylene fabrics but that no measurement or test was capable of predicting the behaviour of all the other materials in dispersal experiments. The inherent variability of dispersal experiments seems to be very great. With a standard deviation of the approximately log-normal distribution of the experimental values as high as about 2 times the mean, it is necessary to carry out as many as 20 replicate experiments in order to differentiate with certainty between garments with a two-fold difference in penetration.

The results presented in this paper confirm the existence of two types of tetracycline resistance in Shigella sonnei. One group of strains had a high level of resistance to tetracycline and oxytetracycline, with a variable level of minocycline resistance. The second group had a lower level of tetracycline resistance and were sensitive to minocycline. After conjugation with E. coli K12 the selected E. coli transconjugants had the same levels of resistance as the parent Sh. sonnei strain, with one exception. Sh. sonnei 87 was resistant to a high level of tetracycline, but was able to transfer only low level resistance. It is suggested that Sh. sonnei 87 carries two plasmids: pSU1, a conjugative plasmid conferring a low level of tetracycline resistance, and pSU2, a non-conjugative plasmid which confers a high level of resistance to tetracycline.

There was a 1000-fold difference between the highest and the lowest mean rate of dispersal of bacteria-carrying particles during exercise among a group of three male and three female subjects. Differences in the numbers of skin fragments dispersed and in the proportion of these carrying bacteria were almost equally responsible for this. Since there is little difference between individuals in the rate of skin replacement the differences in the rates of dispersal of skin fragments during exercise must reflect large variations in the amount of skin surface removed in other ways. The skin fragments dispersed had a wide size range extending below 5 μm for the minimum projected diameter (MPD). The median MPD was about 20 μm with 7–10% less than 10 μm. Many of the particles could therefore pass freely through tightly woven fabrics with pores up to 10–15 μm which might seem to be impenetrable to whole corneocytes, typically larger than 30 × 40 μm in the hydrated state.

The stools of 37 newborn babies born in hospital were examined for the presence of virus. An extract of every stool passed by each baby was examined in the electron microscope and inoculated into cell cultures. The babies were delivered in four separate maternity units (A–D). All the babies from units A and C (9 babies) were found to be excreting rotavirus though none showed any evidence of diarrhoea. Two of the babies also excreted astrovirus. Subsequently unit A was closed for cleaning and, on reopening with more restrictions on visitors, a further 19 babies were examined. No virus was found in any of their stools, nor was virus observed in the stools of babies from units B and D, where visiting was also more restricted. Visiting restrictions in these units excluded older siblings of the babies. No virus was cultured from any stools in this study.

Five methods for the enumeration of micro-organisms in food (pour plate, surface spread plate, surface drop, agar droplet, and microdilution) were used in parallel to examine 100 samples selected from a wide range of food products.

Statistical analyses of the results showed that the regression and correlation coefficients between the methods were highly significant; the lowest correlation coefficient derived for any pair of methods was 0·979. A variation in count between the methods of less than 0·5 log10 cycles was given by 98% of the samples.

Analysis of the operation times and material requirements of each method showed that substantial savings in cost, time, space and support labour were achieved with the microdilution and agar droplet techniques as compared with the conventional pour plate and spread plate methods.

Ten porcine enteroviruses, 2 porcine adenoviruses and 1 coronavirus were isolated directly from 32 samples ofslurry collected from a pig fattening house. Concentration of the same samples by adsorption with the polyelectrolyte PE-60 yielded 24 porcine enteroviruses and 3 porcine adenoviruses. A porcine enterovirus was isolated, following PE-60 concentration, from 1 of 6 slurry samples from a sow farrowing house. No virus was isolated from 12 samples of slurry from dairy cows nor from 6 slurry samples from a calf-rearing unit. A porcine enterovirus was isolated from soil samples, after concentration with PE-60, collected 1, 2 and 8 days after pig slurry was spread on hay stubble. Two porcine enteroviruses were isolated by membrane filtration from 26 samples of surface run-off from land on which pig slurry was routinely spread, and 2 bovine enteroviruses were isolated from cattle feedlot run-off after adsorption to layers of talc and celite followed by hydroextraction. A porcine enterovirus was also isolated from 1 of 33 samples of surface water collected on farms on which pig slurry was routinely spread on the land, but no virus was isolated from 36 samples of ground water from the same farms. The surface water and ground water samples were concentrated by talc-celite adsorption and hydroextraction.

In a study on the serotypes of Escherichia coli which are present in the faeces of students travelling from many parts of the world to Dublin, a great variety of types was found. It was not possible to relate certain types to the various parts of the world where the students came from. There was no decrease in variety of serotypes after the students had been in Dublin for a few months and no distinctive ‘Dublin’ types were acquired. Serological variation of the E. coli was suggested as partially explaining some of the variety of the serotypes found. Although eight students developed diarrhoea without an aetiological agent being isolated after arrival in Dublin, no distinct pattern of E. coli could be related to this.

Roof-collected rainwater is a common water source in subtropical regions and has not been associated with human illness. In Trinidad, the West Indies, a church group, attending a rural camp, developed gastrointestinal illness, caused by Salmonella arechevalata. This rare serotype was isolated from stool specimens of campers, foods eaten at the camp, and a water tap, which was supplied by a storage tank of roof-collected rainwater. The surface of the roof, used as water catchment, was covered with bird faeces. It is postulated that rainwater, falling on the roof, washed off animal excrement which contained S. arechevalata and led to the outbreak of salmonellosis through camper ingestion of contaminated food and water.

We have assessed the effectiveness of flexible-film negative-pressure isolators by physical and biological means. We have found that they afford a high degree of containment and therefore also of safety to hospital staff. We offer some recommendations on the operation of these isolators to ensure the optimum degree of protection.

Studies were made to find evidence of louping-ill virus infection in free-living red grouse and relate this to their breeding success. In areas where ticks were abundant 61 (84%) adult grouse had antibody to the virus compared with 1 (10%) in areas where ticks were relatively scarce. Of 162 chicks tested 25 were shown to be viraemic. Infected chicks were of significantly less weight than comparably aged uninfected birds and the probability that they died was much greater than that of uninfected birds. It is concluded that the relatively poor breeding success in areas of high tick numbers was principally due to infection with louping-ill virus. The susceptibility of the red grouse to infection is discussed.

J. Gallagher and D. M. Horwill, Journal of Hygiene 79, 155–60.

Page 156, paragraph 2, line 4. For 0-0035% read 0-0025%