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.