The intramuscular injection of large amounts, 2–3 g., of Trichomonas foetus crude antigen into five young calves during the first 3 weeks of life produced no antibody.

When the animals were tested with the standard challenge at intervals of several months they responded feebly, showing a definitely impaired capacity to produce antibody, when compared with the response of normal animals.

The ages of the animals still showing this partial paralysis ranged from 1 year and 5 months to 2 years and 2 months.

The epidemiology of Q fever in the residents of two rural areas, respectively, the Romney Marsh in Kent and Chatteris-Witchford rural districts in Cambridgeshire, has been studied.

In the Romney Marsh some 14 % of persons who had suffered from pneumonia or undiagnosed fever probably had Q fever. No cases of Q fever were detected in Chatteris-Witchford R.D.

Two sources of infection—sheep and infected cows' milk—were found in the Romney Marsh, and the infection of some of the cases of Q fever could be attributed to them. There was also a group of cases who were probably not infected from milk or by direct exposure to sheep. They may, however, have been indirectly infected from sheep.

The absence of Q fever in the Chatteris-Witchford area is of interest in view of the use of raw milk by a substantial proportion of the inhabitants of this area and also because of their contact with cattle. Investigation, however, failed to reveal any infection of the cattle with Rickettsia burneti.

No evidence was obtained to support the hypothesis that the observed reduction in numbers of coliform bacteria during sewage treatment was due to the action of bacteriophage.

When examined under similar conditions in the laboratory, the two strains of myxoma virus used to initiate the Australian and European epizootics, and three somewhat attenuated strains recovered in the field in Australia and France, were found to produce lesions containing approximately the same amount of virus, and were transmitted by mosquitoes with approximately equal efficiency. The laboratory variant neuromyxoma multiplied poorly in the skin and was very rarely transmitted by mosquitoes.

The survival potential of different virus strains in the field is correlated with the length of time during which rabbits infected with different strains present to biting mosquitoes lesions containing high concentrations of virus. Sickness of the infected rabbit, and extensive viruliferous skin lesions, will increase its suitability as a source of virus for mosquitoes. These considerations explain the dominance of slightly attenuated strains of myxoma virus in the field in Australia.

Species of mosquitoes, when tested under standard conditions in the laboratory, differ in their efficiency as vectors of myxomatosis. These differences may be related to differences in the method of feeding, and to structural differences in the mouthparts.

Engorgement on an immune rabbit after probing through myxomatous skin lesions had no greater effect on the subsequent ability of mosquitoes to transmit myxomatosis than a similar blood feed on a normal rabbit.

Nine serial mosquito-bite passages of the standard laboratory strain of virus, each involving one week in the rabbit and two weeks in the mosquito, had no effect on its pathogenic behaviour.

Methods are proposed for measuring the constants of an exponentially declining population under conditions in which the least-squares method is unsuitable. A table is given to facilitate the calculations.

1. Elementary body suspensions of vaccinia virus placed on the chorio-allantoic membrane of the developing chick embryo showed immediately an enhanced infectivity which may amount to a several-fold increase in titre.

2. In studying the growth curve of vaccinia virus in the chorio-allantois the distribution of virus between the liquid and the membrane had to be taken into account.

3. Some virus disappeared after inoculation, but this did not necessarily indicate a non-infective phase in a growth cycle.

4. Virus began to increase in 4–5 hr. and proceeded without stepwise increments, reaching a maximum in about 36 hr.

5. Haemagglutinin and complement-fixing ‘soluble’ antigen were formed during the growth of virus, and the titre of each was directly related to the amount of virus. They were detectable when the titre of virus reached respectively 106.5–10.70 and 105.5–106.0 pocks per ml.

6. Haemagglutinin may have been acting as a complement-fixing antigen.

7. Neither haemagglutination nor complement-fixation provided evidence for the occurrence of a non-infective phase of virus as part of a growth cycle.

1. A good immunity was produced in chickens against oral infection with Salm. gallinarum by the use of either of two live attenuated vaccines, one smooth, 9S and one rough, 9R. Dead vaccines had little or no effect.

2. The immunity possessed by chickens vaccinated with 9R at 7 weeks of age commenced to wane 12 weeks later. Chickens similarly vaccinated with 9S were completely immune 4, 12 and 34 weeks later. When tested during the active laying period (at 20 weeks after vaccination) their immunity, although substantial, was not complete.

3. Vaccine 9R did not produce demonstrable agglutinins against smooth Salm. gallinarum in chickens. It was non-lethal to 1-day-old chicks and was never found to revert to the smooth virulent form in vitro or in vivo. By contrast, vaccine 9S produced agglutinins and was lethal for 1-day-old chicks, its virulence not being increased by passage. No deaths were ever observed when this vaccine was used in either chickens of 5–12 weeks old or laying hens.

4. Although both vaccines produced a good immunity when employed in laying hens, vaccination with 9S was accompanied by a marked reduction in egg production; this was not the case with 9R.

5. An immunity of the interference type was also evoked by both live vaccines. This immunity was still produced to some extent in chickens injected with 9S, 2 days after they were infected with Salm. gallinarum by mouth.

6. The interference effect could not be produced by the use of either dead Salm. gallinarum vaccines or certain non-specific substances. The intravenous injection of indian ink had a marked adverse effect on the course of the disease. Injecting chickens with a virulent strain of Bacterium coli at the same time as they were infected with Salm. gallinarum resulted in the latter infection running a milder course than would normally have been expected.

Incomplete antibodies demonstrable by the Coombs antiglobulin technique were found in almost all guinea-pigs inoculated subcutaneously with Brucella abortus. They were detected between the 20th and 30th day after infection and reached a peak between the 50th and 70th day, at which time their titre was 4—8 times that of the ordinary agglutinins.

Paper electrophoresis of sera from infected guinea-pigs showed a great increase in y-globulin as infection developed. Smaller increases in α1- and β-globulin were observed.

Vaccinia virus inoculated on the chorio-allantois and causing infection spread to the uninoculated part of the membrane, the embryo, the allantoic and amniotic fluids and the yolk within 1 day.

The virus grew in the embryo and possibly in the uninoculated part of the chorio-allantois, but in the latter no macroscopic lesions were detected.

Virus increased in amount in the allantoic fluid and may have grown in the lining of the allantoic sac; it did not increase in the amniotic fluid; it was present in yolk and increased slightly but its local growth was doubtful.

The presence of haemagglutinin and complement-fixing antigen was related to the amount of virus.

Allantoic fluid from infected eggs could be a ready source of antigen for complement-fixation.

1. One per cent of samples of farm milk were found to contain pathogenic varieties of Bacterium coli of the O groups associated with infantile gastroenteritis.

2. Unheated cows' milk is a dangerous food for babies.

3. Summer diarrhoea was probably, for the most part, milk-borne infection by special O groups of Bact. coli.

4. Thirteen per cent of chickens on farms were found to be excreting Bact. coli of the O groups associated with infantile gastro-enteritis but only a few of the strains possessed H antigens of varieties known to have an association with human disease.

5. Possible reservoirs of infection cannot be adequately searched for until a selective method is devised to aid the separation of pathogenic O groups of Bact. coli from the non-pathogenic.

1. The immunity to Salm. gallinarum infection produced by injecting chickens with live cultures of different species of Salmonella and other bacteria has been studied.

2. Salm. pullorum produced a complete immunity equal to that produced by Salm. gallinarum itself. The immunity evoked by 11 of 12 other Salmonella species belonging to group D, although substantial, was less complete. Three of four rough group D strains and an avirulent strain of B. coli produced no immunity. Although a considerable degree of immunity was produced by the fourth rough strain it was not equal to that produced by rough strains of Salm. gallinarum.

3. Owing to their lethal nature, it was not possible to be definite about the extent of the immunity produced by two group B strains, Salm. typhi-murium and Salm. Chester, although it may have been equal to that produced by the group D strains.

4. A slight but definite degree of immunity was produced by two group C and two group E Salmonella cultures. Since this immunity was no better than that possessed by chickens previously infected with either Pasteurella septica or Staphylococcus aureus, it was considered to be non-specific in character.

1. Three strains of virus were isolated in suckling mice from the blood of three patients suffering from dengue during a local outbreak of the disease.

2. One strain was adapted to adult mice and identified as a dengue virus of type 1. The identification was confirmed by workers in America.

3. Evidence is presented that the other two strains of virus are similar to the one fully identified.

4. Attention is drawn to the need for detailed work to find a more sensitive method of isolation for dengue viruses.

Of 629 fresh fish (sixty-five species) and forty-two samples of fish-washings obtained from a fish-market in Colombo, and investigated bacteriologically for the presence of Salmonella, thirty-nine fish (twenty-four species) and five samples of washings were found to be harbouring the organisms.

Of fifteen salmonella types found, all except Salm. chittagong and Salm. oranienberg had been previously isolated from human and animal sources in Ceylon.

The presence of Salmonella in the intestines of fresh fish examined suggested that the organisms had entered the live fish, whilst their isolation from the gills and fish-washings was evidence that the fish could have been contaminated post-mortem.

The significance of these findings in relation to hazard to health is discussed.

1. Eleven thousand two hundred and twenty-eight sera from persons, mostly blood donors, living in Kent and seven counties in East Anglia, have been tested and 379 (3·4%) found to have complement-fixing antibody to Rickettsia burneti at a serum titre of 1/10 or greater.

2. A group of 396 donors in Kent and East Anglia, composed of 190 donors with antibody, and 206 donors from those without antibody to serve as controls, were interviewed, and an epidemiological history was obtained on the possible ways in which they had been exposed to R. burneti during the period 1942–53.

3. A comparison of the epidemiological histories of the positive and negative donors showed that donors in Kent who had constant occupational exposure to cattle and sheep including unprocessed materials from them, or who had had a raw milk supply to their homes for all or part of the period 1942–53, more frequently had antibody than those donors without such qualities. In East Anglia the positive serological state among donors was associated only with the use of raw milk and not with occupational exposure to animals. There were suggestive findings among Kentish donors that those who visited farms or other potentially infectious places, either during the course of their job or during their leisure, also experienced greater risk of infection.

4. In general, these findings of infection from occupational exposure, from visits to infectious localities, or from the use of raw milk in the home, confirmed the tentative epidemiological conclusions drawn from a previous investigation of patients with Q fever. It was not found in this survey, however, that residence near dairy farms, receipt of objects from potentially infectious localities, or contact with persons from such localities was associated with a special risk of infection.

5. No association was found between the donors' serological state and the keeping of dogs, cats, poultry, pigs, rabbits, psittacine birds in the vicinity of the home or its infestation by mice, rats or mosquitoes.

The observations reported here show that the T Vi antigen is effective in stimulating antibodies when it is attached to the cells. It is also clear that the Vi antigen is readily removed in aqueous solutions, but in such organic solvents as acetone, chloroform, ether, alcohol, carbon tetrachloride and benzene it is retained by the cell. In alcohol-water mixtures the Vi antigen is progressively more soluble in those solutions where the alcohol concentration is less than 70 %. However, its solubility in water is limited, and by making progressively heavier suspensions it was found that a point is reached at which Vi antigen is retained by the cell, the suspending solution apparently being saturated. Such an aqueous suspension is able to stimulate the production of Vi antibody.

The administration of dried organisms in peanut oil resulted in the production of antibodies against both O and Vi antigens. This was in contradistinction to the results obtained when mineral oil, mineral oil + saline, or peanut oil + saline were used: in these cases the O antibody titre was lower and Vi antibodies were produced in low titre or not at all. It seems clear that the presence of water in the vaccine has the effect of removing the antigenicity of the Vi hapten; water-free, vegetable-oil vaccine is suggested by these studies as an effective agent for immunizing against all of the antigens of Salmonella typhi. The use of such a vaccine in producing active and passive immunity in animals is discussed in the paper which follows.

The complement-fixing antigen activity of canine hepatitis virus is due to particles which are estimated to measure from 70 to 105 mμ and which probably can form larger aggregates. The active particles in the original virus tissue suspension were of sizes similar to those in the purified preparations, their serological properties being identical. The chemical properties of the antigen are those of protein, probably ribonucleoprotein. No evidence has been obtained to explain the association of a high concentration of the antigen with the presence of numerous intranuclear inclusion bodies in the lesions. It is evident, however, from these experiments that the inclusion bodies are not just deposits of the serologically active material. Since the intranuclear inclusion bodies and also the antigen are probably composed of ribonucleoprotein, it is possible that they both represent certain stages in the formation of virus protein by a catalytic reaction from the denatured host proteinogen.