Several methods were used to obtain serologically active materials from cultures of Micropolyspora faeni. From the results of immunodiffusion and immunoelectrophoresis tests on these materials it is suggested that preparations for the laboratory diagnosis of farmer's lung disease (FLD) should contain concentrated culture supernatant (CS) and extracts of mycelium obtained by ultrasonic treatment (MU). Although CS and MU have many serological activities in common they also possess activities unique to each.

Extraction of mycelium with trichloracetic acid, boiling water or methanol yielded a product which gave simple patterns in immunodiffusion tests. The products contained little protein but were rich in carbohydrates, particularly arabinose, galactose and glucosamine. A similar material was obtained from a cell-wall preparation by treatment with lysozyme. Antibodies to the serologically active substances in these materials occurred more frequently in sera of patients with FLD than antibodies to any other M. faeni antigen.

Attempts to obtain serologically active materials from spores were unsuccessful. Moreover antibodies to M. faeni could not be removed from patients' sera by absorption with partially purified spore preparations. It is suggested that the hypersensitivity in FLD arises from exposure to mycelial antigens.

The indirect immunofluorescent technique has been used to study the specific immunoglobulin responses in nasal secretions from ten adults with acute rubella. Titres of IgA antibody in nasal washings usually exceeded those of IgG, but both types of antibody were detected in all patients. They appeared a few days after the rash, reached maximum titres during the second week and then declined. IgA antibody was no longer detectable after 47 days and was not detected at all in nasal washings from adults who had experienced rubella in the past. Low titres of IgG antibody persisted in some patients for longer than IgA and traces of IgG were found in nasal washings from a minority of adults with a past history of rubella. Nasal antibodies in acute rubella are therefore transient and unlikely to take part in resistance to reinfection.

In sucrose-density gradients nasal IgA antibody sedimented more rapidly than IgG and there was little overlap between these two types of antibody. IgA antibody in serum was more heterogeneous; it was found in nearly all the fractions which contained IgG antibody and in many of those which contained IgM.

Low agglutinin titres to pertussis suspensions were found in 99% of sera from a group comprising healthy adults and non-vaccinated, non-infected infants of 1–6 months of age. These are attributable to agglutinins to heat-stable antigens and/or heat labile agglutinogen 1, and cross-absorption tests must be done on the sera in order to distinguish between the two. Agglutinins to agglutinogens 2 and 3 were found in only about 20% of adult sera. Bactericidal antibody was low in titre or absent in all sera from non-exposed individuals.

Raised bactericidal antibody titres and the presence of agglutinins 2 and 3 were attributed to exposure to Bordetella pertussis antigens, either as vaccine or as infection. The variation, amongst both vaccinated and infected children, was very great. A vaccinated child who became ill responded to the infection in much the same way as a non-vaccinated child. We were unable to relate the immunity of the child to the titres either of agglutinins or of the bactericidal antibody.

The protective ability of sera from vaccinated or infected children measured in mice against small, lethal brain infections was also unrelated to the state of immunity in the children, but this protective ability was correlated with the complement-mediated bactericidal antibody titres of the sera.

The distribution of agglutinins, bactericidal antibody, and anti-haemagglutinin in serum IgG and IgM was different in vaccinated and infected children.

A total of 2716 R factors and transfer factors isolated from Escherichia coli and salmonellas of human and animal origin were studied for their phage-restrictive effects in Salmonella typhimurium phage type 36. All of 1402 wild fi+ factors were non-restricting. The F factor of E. coli K12 was unique among the F-like factors tested in that it inhibited lysis of type 36 by one typing phage. In contrast, eleven distinct changes in the phage type of 36 were produced by fi− I-like factors. I-like plasmids can thus be subdivided by this method.

I-like R factors and transfer factors from human and animal enterobacteria were categorized by their phage-restrictive effects in type 36. Factors resembling Δ in this respect predominated among fi− I-like factor from human E. coli and S. typhimurium and from porcine E. coli. Δ-like and Coll-like fi− factors were equally distributed in bovine S. typhimurium. Coll-like factors were commonest in bovine and avian E. coli.

The antibiotic resistance patterns of coliforms in faecal specimens from pigs and their human contacts were studied. The ability of the resistant coliforms to transfer their resistance in vitro to antibiotic-sensitive recipients was examined. The results showed that pigs which had received antibiotics carried more multiply-resistant, R-factor bearing coliforms than pigs which had not been given antibiotics. Human contacts of the antibiotic-treated pigs had a higher incidence of antibiotic-resistant coliforms with R-factors than human contacts of pigs which had not been given antibiotics. It is concluded that antibiotic treatment of farm, animals may lead to acquisition of antibiotic resistance by gut coliforms of man.