The success of detecting Mycobacterium tuberculosis in milk by microscopic examination of samples from each quarter of the udder and the use of this method in rapidly eliminating from a herd those cows which were excreting tubercie bacilli has been reported (Maitland, 1937). The advantages from the clinical point of view with illustrative clinical data have been recorded by Locke (1937), who collaborated in the investigation.

The value of ‘Statistical control’ in haematology has been demonstrated. After a short historical survey of the chief contributions to the theory of the technique of counting and of the application of the Poisson distribution in biology, the statistical theory of the distribution of cells in the haemocytometer is briefly described and examples are given of large sample methods of testing goodness of fit. Some of the author's counts are considered.

In the red cell counts a ‘crowding’ effect is noted by which the variance between the counts on the individual squares on the haemocytometer is reduced. This effect is trivial in white cell counting owing to the small proportion of the area occupied, but is important in reducing the variance between the numbers of red cells on individual small haemocytometer squares. A quadratic function adequately describes the regression of the variance on the mean for fixed size of haemocytometer square, for the usual range of red cell density. This function, calculated on the basis of the author's own counts also describes Berkson's findings on crowding adequately and is not, considering the approximations involved, inconsistent with the theoretical discussion on crowding. Reasons are given for believing that the crowding effect is only trivial if blocks of 16 squares are compared in red cell counting.

The methods of statistical control, suggested by R. A. Fisher's work, are introduced. It is shown that they give satisfactory results when applied to some of the author's own counts. The methods are used to discuss the consistency of the counts of certain medical graduates and technicians. The methods are suitable for a review of the literature with the object of examining the internal consistency of the counts made by various authors.

Since the introduction of toxoid-antitoxin floccules by Glenny & Pope (1927) little has been done to improve their value as an antigen. Watson, Taggart & Shaw (1941) and Barr (1949) confirmed the work of these investigators in showing that heating at 80° C. increased their antigenic value.

It has been shown in this article that about 80% of the toxoid contained in floccules may be recovered if they are dissolved in n/20-NaOH, the alkali being allowed to act for 1 hr. before being neutralized with acid. A purity of between 0·002 and 0·0025 mg. N/Lf is obtained regularly. Such a dissolved floccule solution is a relatively poor antigen when tested in guinea-pigs, but may be rendered highly antigenic when precipitated with alum or adsorbed on AIPO4.

The method has the advantage of simplicity. It would appear that any medium capable of yielding a toxin convertible into flucculable toxoid may be used, thus doing away with cumbersome media-making methods. However, for the best results in terms of both recovery and purity, a high-value, quickly flocculating, toxoid should be used. Further, the choice of antitoxin appears to be important where recovery is concerned. With most of the antitoxins used, recoveries of between 70 and 90 % were got, although with that used, for most of the work recorded here, the recoveries were much lower.

It is not yet established which is the antigen of choice—alum-precipitated (P.D.F.) or AIPO4-adsorbed (A.D.F.) dissolved floccules. A.D.F. has the advantage of ease of preparation and of constant mineral carrier content, but further work is necessary to establish the optimum amount of carrier before a definite answer can be given. A paper on immunity responses in man will be published shortly in the Lancet.

When mussels are allowed to function in sea water the main changes occurring in the water are depletion of dissolved oxygen and lowering of pH value. Provided the faeces and pseudo-faeces are not disturbed the increase in the content of organic matter is not appreciable. If the supernatant water is removed and aerated with diffused air the water is re-oxygenated, the pH value is restored to its original level, and the water so treated may be re-used for immersing a fresh batch of mussels. The process of re-use may apparently be continued indefinitely.

Under these conditions a high degree of cleansing is achieved by the mussels, the count of coli-aerogenes bacteria being reduced to a small fraction of the original; counts of bacteria in the water, on the other hand, are subject to large fluctuations. It was considered advisable for this reason to chlorinate the water between each cycle of cleansing in order to immerse the mussels on each occasion in water which itself was reasonably certain to be free from coliform bacteria, potentially including pathogens.

Experience showed that the greatest difficulty involved in the use of chlorine was in removing residual chlorine so that it would not inhibit functioning of the mussels when the water was re-used. After trials of various alternatives it was concluded that the most satisfactory method was to determine by means of small-scale tests the smallest quantity of chlorine (below the break-point) required to give a residual concentration of 0·05–0·10 p.p.m. in the re-used water and then to add the corresponding quantity with precision to the bulk of water by means of a dosing apparatus while the water was being pumped from the mussel tank to the aeration tank. After a period of contact of 1 hr. the water was aerated for a further hour. A series of trials in a semll-scale plant showed that this treatment ensured that residual chlorine in the water being added to the mussels did not exceed 0·05 p.p.m.; as a result the mussels functioned satisfactorily and the degree of cleansing attained was comparable with that attained in the existing mussel-cleansing tanks in which sterilized fresh sea water is used for each cycle of cleansing. Although the concentration of residual chlorine was small the mussels kept the water so clear that this concentration was effectively bactericidal and the bacterial quality of the water was usually comparable with that of good drinking water. Low temperatures retard the metabolic activity of mussels and below 4° C. this is so marked that the degree of cleansing achieved is unsatisfactory. For this reason it is recommended that the temperature of the re-used water should be maintained at 6° C. (43° F.) or above.

Oysters were found to be satisfactorily cleansed by a process of re-use similar to that adopted for mussels, provided the temperature of the water was maintained at 54° F. (12·2° C.) or rather higher.

Small-scale trials showed that artificial sea water, prepared by dissolving in fresh water suitable quantities of the major constituents of natural sea water, could be successfully re-used for cleansing mussels.

1. Repeated inoculation of thirty-six guinea-pig does with random blood of guinea-pig bucks failed to produce any demonstrable anti-red cell isoantibodies.

2. A note is included on the blood picture of a small series of newborn guinea-pigs aged between 12 and 24 hr.

A survey of the extent of bovine type human tuberculosis in Northern Ireland has been made from unselected material during the past 5 years.

Only one case of bovine type pulmonary tuberculosis was found out of more than 1000 cases.

The proportion of bovine type cases in the non-respiratory group was 4·3%; 3·3% of meningitis cases were of bovine origin.

A marked decline in the extent of bovine type infection is shown to have occurred during the past 10 years. This is believed to be due to a great increase in the pasteurization of milk during this period. Milk was a particularly valuable food during the war, and pasteurization facilitated its distribution.

1. An account is given of the relationship of Bact. coli type O55 B5 H6 to cases of infantile gastro-enteritis.

2. In Aberdeen this organism has now largely replaced the O111 variety (alpha) which had been the prevalent type during 1947 and 1948 in its association with the disease.

3. A brief description has been given of the clinical findings, which show no marked deviation from those found in similar cases elsewhere. The actual severity of the illness as judged by the mortality rate has been greatly lessened.

4. The value of medication with chloromycetin in a small number of cases has been discussed. It is believed that the use of the drug eliminates the organism from the bowel in certain cases and helps to reduce the number of cross-infections even in a cubicle ward. The average weekly weight increase of children on chloromycetin was almost double that of a group of healthy, untreated children.

5. Several small outbreaks of infantile diarrhoea associated with the O55 B5 H6 type in wards and institutions have been recorded. In some the numbers of symtomless carriers of the O55 type have exceeded those of clinical cases also harbouring the same organism.

6. ‘O’ and ‘H’ agglutination tests made with suspensions of the prevalent strain of Bact. coli on sera from cases examined early in the disease were mostly negative. Such positive titres as were obtained were low. Examination of paired acute and convalescent sera from 29 cases gave some evidence of increasing titre during the course of the disease in eight cases, but such increase was of low order.

7. The present bacteriological evidence suggests that further specific types of Bact. coli will be found to be associated with the disease.

Techniques for the fractionation of C′ 1 and C′ 2, and for the specific inactivation of C′ 4 of horse complement are described, and shown to be satisfactory. These three components of complement and also conglutinin are found to be essential to the process of conglutination.

The experiments reported do not exclude the possibility of additional or unidentified fractions of horse serum playing a part in the reaction. Whether or not C′ 3 is essential to the process of conglutination could not be determined from the evidence available.

It was found that by the procedures employed the C′ 4 fraction may be supplied either from the horse complement or from the heated bovine serum added as a source of conglutinin or from both.

When horse complement and bovine antibody against sheep cells are used it is found that conglutination will only result when the complement components are absorbed on to the immune complex in a particular sequence. The sequence is that C′ 1 is absorbed first on to the sensitized cells, and then C′ 2 and C′ 4 are absorbed together. Both these components must be presented together to the sensitized cells carrying C′ 1 if conglutination is to result. Finally, conglutinin acts on the sensitized cells which have absorbed the three complement components, and conglutination results.

The significance of these findings, together with matters of a more general nature, is discussed.

Filtrates prepared from suspensions and broth cultures of Brucella vaŕiants were examined for the presence of Brucella bacteriophage. Smooth Brucella, when mixed with decimal dilutions of these filtrates in broth, (1) contained typical bacteriophage plaques when inoculated on solid medium, and (2) gave a heavy turbidity in broth after suitable incubation, but frequently failed to yield visible growth upon plating these broth cultures to a solid medium. These data were discussed in terms of Brucella bacteriophage, and it is concluded that the presence of this virus best accounted for the phenomena observed.

To estimate the number of viable streptococci present in the air of occupied places, airborne bacteria were collected in a slit-sampler on a nutrient agar medium containing 5% sucrose, 5% horse serum, 0·25 mg./100 ml. crystal violet, and 1·0/100 ml. potassium tellurite. This medium inhibits the great majority of the staphylococci and micrococci found in the air. On it colonies of Strep. salivarius are readily recognized by their mucoid form. The number of other streptococci was estimated by picking and examining a random sample of the colonies.

Streptococci were recognized in the sample by three different methods: (1) by microscopic examination of a nigrosin film made from the original colony; (2) by examination of a nigrosin film and also subculturing the colony to a ditch plate having blood agar on one side and serum agar containing 40% or bile on the other; or (3) by subculture to bile-agar ditch plates alone, the bile agar in this case containing aesculin and ferric citrate. By method (1) streptococci could be distinguished from micrococci; by methods (2) and (3) this distinction could be made with more confidence, and in addition enterococci could be distinguished from other streptococci.

A number of tests of the efficiency of the method were carried out, and it was concluded that some 20–40% of the viable streptococci in the air were missed.

The standard errors of the mean of a number of streptococcal counts in schoolrooms were calculated and found to be of the order of 14% of the mean when our standard routine was followed; increasing the size of the sample of colonies picked for examination by 30% would probably have had only a trivial effect on the standard error of the mean, since the greater part of the variation was due to differences between the counts in different rooms.