Normal virulent B. anthracis all possess the genetical potentiality of producing capsules, but the capsule is only produced in vivo or under special cultural conditions suitable for capsule formation, especially in the presence of CO2.

During multiplication the potentially capsulated cells regularly produce a small number of a stable non-capsulated avirulent variant. All laboratory stock cultures of B. anthracis examined were found to contain from 0·14 to 32·4% of this avirulent variant. The variant has also been demonstrated in hair and wool from anthrax endemic areas, which suggests that the dissociation may also occur in nature.

Both the normal potentially capsulated cell and the variant incapable of producing capsules grow on plain agar in air as rough colonies with apparently non-capsulated cells. But when they are grown on serum agar in the presence of CO2, the normal potentially capsulated cell gives rise to a smooth mucoid colony with fully capsulated cells, while the variant still produces a rough colony with non-capsulated cells. If we regard the smooth mucoid colony produced by normal virulent B. anthracis on serum agar in CO2 as the normal colony form of B. anthracis, then the dissociation of the avirulent non-capsulated variant with the rough colony form falls in line with the usual S-R dissociation of other bacteria.

The more widely known variation of B. anthracis, which had long been held as an exception to the usual S-R variation of other bacteria, refers to colonies grown on plain agar in air. This variation is different from the one just described and is not so clear-cut and significant. Its underlying cause is not quite clear except that the chain length of the growth appears to play some part.

A summarized picture showing the whole range of variation in colony appearance and cell structure of B. anthracis is presented.

1. Samples of whey from individual cows and herds infected with R. burneti and from uninfected cows and herds have been tested for R. burneti antibodies.

2. The haemolytic complement-fixation test, the conglutinating complement-absorption test, the direct agglutination test and the anti-globulin sensitization test were compared for sensitivity in detecting whey antibody.

3. The direct agglutination test was generally more sensitive than either of the complement-fixation tests, particularly when it was enhanced by centrifugation of the sensitized rickettsiae. The anti-globulin sensitization test, however, gave the highest titres andit appeared to be specific.

4. The practical application of these techniques and certain theoretical considerations are discussed.

By a combination of a swab sampling method, modern cultural techniques and Vi-phage typing, a survey for enteric organisms has been made of the sewerage system of a town of about 10,000 inhabitants.

The location of typhoid and paratyphoid B carriers by systematic swabbing of sewers in accordance with cultural findings has been proved feasible.

Two paratyphoid B carriers and one typhoid carrier were discovered by this method. Evidence was obtained of the presence of at least six other foci of paratyphoid infection in the town.

Paratyphoid bacilh were repeatedly isolated from a river flowing through the town. The possible role of river pollution in the maintenance of endemic enteric infection in the area is discussed.

The methods used in this survey might be applied with advantage to the elucidation of unsolved enteric outbreaks, to the control of food-handling establishments and to epidemiological studies of other infections in which the causative organism gains access to sewage.

We are indebted to Mr R. W. Pinney and Mr W. E. Coles, who did by far the greater part of the work of laying the sewage swabs and sending them to the laboratory; to Mr L. M. Blanchard, Surveyor, Sidmouth U.D.C., for his assistance with the preparation of maps of the area; to Dr A. Felix for his kindness in typing the enteric strains isolated; and to Mr R. N. Brown and Mr S. Barlow for much assistance in the development of cultural techniques.

Specific precipitating antisera can be prepared in rabbits by the repeated inoculation of alum-precipitated serum.

High titre of sensitivity to the homologous serum is obtained only with a complete loss of specificity.

By absorption of the non-specific antiserum with heterologous sera the specificity can be restored.

Two types of reactions can be observed in such non-specific antisera: (1) the homologous reaction which occurs with the sera of closely related animals, and (2) the heterologous reaction which is attributed to the presence of antigens which are widespread among mammalian sera.

Cross-immunization experiments have demonstrated the presence of specific antigenic material in the serum of indvidual species.

A schematic representation of the antigenic arrangement of mammalian sera is suggested.

I am indebted to the Colonial Medical Research Committee for their help in this investigation and for the financial help provided by the Colonial Office.

I would like to acknowledge my indebtedness to Prof. P. A. Buxton, F.R.S., for his help and encouragement and in providing and housing the chimpanzee used in these experiments, to Dr C. H. N. Jackson and Mr W. A. Hilton for their valued help in the field in collecting a number of sera from East Africa, and also to Dr E. Hindle, F.R.S., for supplying me with materials from animals which had died at the Zoological Gardens, London. I wish to record my thanks to Mr E. Pope, for his valued technical assistance throughout these experiments.

1. A total of ninety-nine strains of anaerobic cocci from human beings have been studied and six groups demarcated on the basis of gas formation and fermentation reactions.

2.Gas formation by strains of group I is due to fermentation of glucose, laevu-lose or maltose, with a sulphur compound as an activator and with the production of gas rich in CO2.

3.Gas formation by strains of groups II, III and V is due to fermentation of different organic acids, sulphur not being required, and accompanied by the production of gases containing a high proportion of H2.

4.Two groups, IV and VI, do not from gas, but the latter has marked fermentation abilities.

There is correlation between microscopic apperances, probable pathogenicity, sensitivity to antibiotics and sulphonamides and the groups demarcated.

Ninety-two out of ninety-nine strains isolated from human beings could be placed in one or other group.

We are greatly indebted to Dr H. J. Parish of the Wellcome Research Laboratories for a sample of Polymyxin E, to Prof. L. Young for much assistance and advice and to those colleagues, particularly Dr R. M. Caiman of Queen Charlotte's Hospital, who sent us strains.

Preliminary observations in a systematic study of fifteen bacteriophages isolated from Pseudomonas aeruginosa are described. The evidence as to size, morphology and plaque appearance indicates that this group of phages comprises several distinct types. The electron microscope reveals some of the phages to be simple spheres, but the majority are tailed particles.

The importance of this group in the study of the phenomenon of lysogenicity is pointed out.

1. Treatment of washed sheep erythrocytes with dilute solutions of tannic acid rendered them capable of adsorbing diphtheria toxin from solution. Diphtheria toxoid was not adsorbed under similar conditions.

2. Cells thus sensitized with diphtheria toxin were agglutinable by a proportion of children's sera and human cord sera; sera of rabbits, mice and horses, immunized with soluble or adsorbed diphtheria toxoid, also agglutinated sensitized cells, as did also the sera of children suffering from diphtheria and treated with antitoxic serum. Clear-cut results were usually obtained, but the end-point of titrations on horse sera was, in some cases, difficult to determine.

3. Using intradermal toxin neutralization tests, a highly significant linear relationship was demonstrated between the haemagglutinating titres of children's sera for the sensitized red cells, and the antitoxin content of their sera. The antitoxin content corresponding to the smallest detectable amount of haemagglutinating antibody was 0·008 unit per ml. serum.

4. The haemagglutinating titre was an accurate measure of the mean antitoxin level of a group of sera and a close measure of the antitoxin content of most specimens in individual tests.

Methods of graduation of a series of observations by means of moving averages were discussed by Sheppard (1914), and subsequently by Sherriff (1920) and a number of other writers. These methods based on least squares or weighted least squares solutions differ from actuarial or summation methods. Thompson (1947) has proposed that the method of moving averages be considered a ‘basic’ one in the estimation of the median effective dose (LD50) of bioassay data. On the basis of an empirical study of the data of Topley and Wilson he recommended in particular the use of a three-term moving average. In a recent paper, Finney (1950) has discussed the efficiency of Thompson's moving average method generally.

An epidemic of epidemic hepatitis occurred in the district of a rural Central School in Yates and Ontario Counties of Central New York. The epidemic comprised 176 cases of which 135 were jaundiced, and extended from September 1949 to May 1950, with most cases in March 1950.

No common source could be demonstrated and the transmission was believed to be due to human contact alone.

The overall morbidity rate was about 5%, attack rate among school children 22%, and among the adults and pre-school family contacts of infected school children the attack rate was 18%, with no significant age difference in attack rate.

Differences in attack rate in the various school grades were explained by higher number of siblings of children in certain grades.

Household attack rates among non-school persons were related to number of sick school children, family size, and hygienic standard of the home. The first of these factors seemed most important.

More jaundiced cases occurred in families where the exposure was heaviest.

The sensitivity of different culture media for isolating salmonellae from the faeces of man, dog, horse, cow, sheep, pig, chicken, duck and turkey has been investigated. Suitable dilutions of tissue fluids of animals that had died from salmonella infection, and in which the numbers of viable salmonellae could be accurately estimated, had previously been added to the faecal specimens. Tissue fluids containing Salmonella dublin, typhi-murium, thompson, cholerae-suis, gallinarum and pullorum were used.

Selenite and tetrathionate media were greatly superior to liquid desoxycholatecitrate medium, liquid Wilson and Blair medium, cacotheline broth and brilliant green peptone water. By the use of either selenite or tetrathionate media it was usually possible to recover salmonellae from faecal specimens to which less than ten salmonellae had been added. They could nearly always be recovered from specimens containing 100 salmonellae.

Salmonellae were more easily recovered from the faeces of some species of animals than others. For selenite, the order of ease of recovery was horse, followed by sheep, human, cow, chicken, pig, turkey, dog and duck faeces, and for tetrathionate, horse, sheep, human, dog, pig, cow, turkey, chicken and duck faeces.

Selenite medium was preferable to tetrathionate for examining cow and chicken faeces but the reverse was true in the case of dog faeces; slight differences only were noted in the other species. Taken as a whole, selenite was slightly superior to tetrathionate, butbest results were obtained by the use of both media.

Some types of salmonellae were easier to recover from faeces than others.Of six salmonella types, Salm. thompson was easiest to recover followed by Salm. typhi-murium, dublin, gallinarum, pullorum and cholerae-suis, in that order.

Erroneous results were obtained when laboratory cultures were used instead of infected tissue fluids.

It was necessary to add several thousand salmonellae to faeces before they could be recovered by direct culture on desoxycholate-citrate-agar or Wilson and Blair solid medium.

Salm. cholerae-suis and Salm. abortus-ovis were exceptional in that direct plating was superior to the use of enrichment media.

With lightly infected specimens an incubation period of 24–30 hr. was optimum for selenite and tetrathionate media. A longer period was detrimental with tetrathionate but not with selenite medium.

The combination of selenite and solid Wilson and Blair medium was sometimes too inhibitory to permit the growth of salmonellae.

There was little advantage in preparing selenite medium with mannitol instead of lactose.

1. The kinds of bacteria found inside hams and bacon gammons are listed and discussed.

2. Taintinǵ of lightly-or un-injected hams is caused by a flora composed chiefly of various faecal Clostridia and/or faecal Streptococci, which are probably intrinsic to the animal's body. In bacon gammons which are heavily injected with brine, the characteristic flora consists of psychrophilic halophiles, mostly Micrococci, probably derived mainly from the brine; if injection is only partly effective, the intrinsic flora may develop also, faecal Streptococci being the type most likely to survive.

3. Within sound hams, bacteria are virtually absent; in tainting, the numbers may become enormous even before curing is finished. Injected gammons, on the contrary, generally contain bacteria in moderate numbers, but they multiply relatively slowly.

4. In hams, the taint is always confined within a group of muscles near the bone which are less acid, and initially less salty, than the outer muscles. In gammons, off-flavours caused by bacteria may occasionally develop in the outer layers as well.

5. A ham properly cured without injection, and internally ‘ sterile’, can be made to keep almost indefinitely. The bacteria injected into gammons slowly cause internal off-flavours even in cool storage. For long storage, it would seem wise to sterilize brine used for injection, and the addition of acid might be beneficial.

The routine methods for propagation of staphylococcal typing bacteriophages, and for their use in identifying strains of staphylococci, are described.

Most of the phages can be propagated in fluid media as well as on agar, and for some glucose-peptone-water is a better medium than nutrient or Todd-Hewitt broth.

Many phage filtrates derived from broth or agar propagation contain, in addition to the phage, an agent that inhibits the growth of staphylococci.

Investigations of variations in the routine typing technique showed that considerable latitude is permissible in the age of cultures used for typing and in the inoculation procedures. It is, however, important to test all phage filtrates after propagation for purity and freedom from the inhibitory agent; and to repeat the tests frequently during use to detect alterations in titre.

About 40% of staphylococci are not lysed by the phages used at their test dilution; about half of these untypable strains are lysed by undiluted phage filtrates.

An analysis was made of the results of typing 567 independent strains of staphylococci; 229 of these showed strong lysis by one or more phages, but there were no fewer than 82 distinct phage patterns represented, and only 118 strains were lysed strongly by a single phage. Certain phages tend to appear together in patterns and on the basis of such associations three main phage groups can be distinguished; they are known respectively as the ‘3A’, ‘6/47’ and ‘52’ groups.

No method was discovered for segregating patterns into ‘types’, but conventions have been devised on the basis of the variation observed in sets of strains from various sources, for distinguishing between different patterns.

We are deeply indebted to Dr V. D. Allison for teaching us the bacteriophage typing methods and for giving us the benefit of his wide experience.

We are grateful to the following for sending us phages, and in some cases discussing their results with us: Dr Phyllis Rountree, Sydney; Dr R. Wahl, Paris; Dr G. Wallmark, Stockholm; Dr H. Williams Smith, Poultry Research Station, and Mr A. M. Hood, Birmingham.

Our thanks are also due to Miss S. Mayo, Mrs E. Lyons and Mr D. Woodroof for technical assistance. The photographs for PI. 14 were taken by Mr W. Clifford.

Intradermal inoculation of the tongue, subcutaneous, intracutaneous and intra-venous inoculation were compared in determining the relative facility with which foot-and-mouth disease infection can be produced in cattle.

In seven experiments using six virus strains, least virus was required to infect by the intradermal tongue route, but wide variation was observed between the amount of virus required to infect by this route and by the subcutaneous, intracutaneous or intravenous routes. The smallest difference between tongue and subcutaneous inoculation was not significant, and the largest difference was that 250,000 times more virus was required to infect by the subcutaneous route. The readiness with which infection is produced by routes other than intradermal inoculation of the tongue may be greatly influenced by the strain of virus used.

Intranasal instillation of the virus was also compared with intradermal tongue inoculation. As with subcutaneous inoculation, more virus was required for infection by the intranasal route and, similarly, the results suggest that considerable variation would be found in the amount of different strains required.

A correlation is suggested between this variation and the variation observed in the ‘invasiveness’ of strains in cattle exposed to contact infection.

Attention is drawn to the significance of the observations recorded in this paper in relation to tests of foot-and-mouth disease vaccines for non-infectivity.

1. ACTH, cortisone, sphingomyelin and sphingosine produce a delayed in-hibition of the skin reaction in cattle sensitized to Trichomonas foetus antigen when they are injected intramuscularly. The inhibition is greatest 6–9 hr. after injection of the inhibitor, but may sometimes be further delayed. This reaction is associated with an absolute lymphopoenia of slightly variable duration.

N-acetylsphingosine retains the desensitizing effect of crude sphingomyelin and is of much the same potency as cortisone acetate of Merck.

2. The lymphopoenia associated with the injection of ACTH and cortisone occurs also with sphingosine, and these three substances are also alike in not producing the desensitizing effect until some hours after the injection.

3. Phenergan, a synthetic anti-histamine, has a different action, and its desensitizing effect which appears to depend on a direct antagonizing of the oedema-provoking substance resulting from the junction of antibody and antigen (hapten) takes effect apparently as soon as the injected drug itself is present at the site of the injection of the specific testing fluid.

4. Methods of testing the capacity of the hormones and drugs to prevent the skin reaction locally and to nullify the action of the in vitro antigen-antibody mixtures which can produce a reaction in normal animals are described.

Phenergan alone of the substances dealt with has any power to act locally. Phenergan mixed with the antigen-antiserum complex in vitro prevents the formation of the characteristic swelling in a normal animal.

Phenergan mixed with the specific testing fluid prevents the development of the skin reaction when the mixture is injected intradermally into a sensitive animal.

Cortisone and sphingosine do not have this action.

We are indebted to the Medical Research Council for supplies of ACTH and cortisone and to Sir Charles Harington for N-acetylsphingosine.

The personal encouragement of Prof. H. G. Lamont and the support afforded by the Ministry of Agriculture, Northern Ireland, are gratefully acknowledged.

1. Mice infected intracerebrally with M. tuberculosis H 37 Rv in doses of 400,000–1,300,000 bacillary units gave a graduated series of responses whose regression provided an arbitrary 100% standard of mouse virulence for human tubercle bacilli.

2. Under certain experimental conditions this standard remained stable and its confidence limits could be defined.

3. The relative virulence of other strains of tubercle bacilli could be assessed quantitatively in relation to this standard, and expressed as a (percentage) virulence ratio.

4. 100% virulence was defined as that of a strain showing a threshold of response of 600,000–1,300,000 bacillary units, lethal cerebral lesions at 8–20 days or necrotic pulmonary lesions at 20–30 days, and a mean 50% mortality time of less than 36 days.

5. Qualitative and quantitative responses characteristic of intermediate and very low virulence were also defined.

6. Higher proportions of virulent cells were obtained from 3 to 7 days than from older cultures in Tween-albumin liquid media, but inocula were more suitably standardized by obtaining cells from 11 to 12-day cultures.

7. Passage in the mouse and guinea-pig failed to alter the virulence of H 37 Rv. A slight increase was effected by passage in the lung of the rabbit.

8. The technique employed was a measure essentially of the ability of the organism to multiply in, and damage, resistant tissues.

A technique is developed by which minimal changes in haemagglutinin inhibitors can be detected.

It is shown that the inhibitor index of normal human nasal secretions is 33·2 with a standard deviation of ±7·4. This value is independent of the donor, of the absolute amount of mucus collected, of the time of extraction, and of the period of storage.

Minimal amounts of influenza viruses (less than one agglutinating dose) are capable of causing significant alteration in the inhibitor index upon interaction in vitro for 60 min. at 37° C.

An appreciation of the chance of failing to detect small amounts of infectivity is the main consideration in the design of any test for non-infectivity. The absence of infectivity can never be proved, and many samples must be tested and found negative before there is reasonable assurance that a supposedly non-infective vaccine is safe to use.

The highest percentage, p, of samples likely to be infective for various numbers of negative observations, n, has been calculated from the expression (1−p)n = P for different levels of P, the chance of failure to detect a positive. For example, if 100 observations are made and found negative the percentage of infective samples would be unlikely (P = 0·05) to exceed 3·0 and highly unlikely (P = 0·01) to exceed 4·5. It is suggested that at least this number of observations be made.

The particular problem of detecting foot-and-mouth disease virus by inoculation of a susceptible host is considered in relation to the theoretical requirements of a test of sound design.

The multiple intradermal tongue inoculation of a number of cattle is the method that best meets these requirements. It is shown that this method is the most sensitive for the detection of foot-and-mouth disease virus, at least twenty observations can be provided by one animal, the presence of non-infective but antigenic material in the inoculum does not hinder the detection of a trace of active virus, and an estimate of the initial susceptibility of the tongue tissue of non-reactors can usually be obtained.

Although a much larger volume can be tested by subcutaneous inoculation, disadvantages of this route are shown to be that with some strains very large amounts of virus are required for infection, only one observation is provided by each animal, the presence of non-infective but antigenic material in the inoculum may mask the presence of active virus, and non-reactors to an immunizing dose of vaccine cannot subsequently be tested for a measure of their initial susceptibility relevant to this route of inoculation.

I am indebted to Dr J. O. Irwin, of the Statistical Research Unit, London School of Hygiene and Tropical Medicine, for advice on the use of the binomial distribution and for his helpful criticism of the manuscript; to Dr Ian A. Galloway, Director of this Institute for his interest and encouragement and to Mr W. J. Brownsea for technical assistance during the work described in this and the previous paper.

Changes in the inhibitory properties of human nasal mucus were found to be positively correlated with objective signs of influenza virus infection during the 1950 epidemic. In approximately 80% of the cases the inhibitor index was significantly increased during the acute stage of the disease, and reverted to normal during convalescence.

Various respiratory diseases other than epidemic influenza did not cause similar changes.

Examination of the inhibitory pattern of nasal mucus is proposed as a presumptive test for influenza virus infection, and its merits are discussed under the headings of reproducibility, specificity and sensitivity.

Comparable groups of adult African women were vaccinated (a) with a mixed 17 D yellow fever-vaccinia vaccine administered by scarification, or (b) by inoculation subcutaneously with 17 D vaccine followed by scarification with vaccinia vaccine. There was no significant difference in the response of the two groups to the vaccinia vaccine. In the group inoculated with the mixed vaccine, of those whose sera contained no demonstrable yellow fever antibody prior to vaccination, 66·6% had developed antibody when their sera were tested 28 days later. In the other group 100% had developed antibody by the twenty-eighth day after vaccination.

It is suggested that the difference in the response of the groups in this study might be due to some local interference which prevented invasion by the 17D virus in some cases.

While there is good evidence for the efficiency of 17D vaccine as an immunizing agent when administered by scarification (Hahn, 1951; Dick, 1952), the present study indicates that the percentage of those who became immune after vaccination with the mixed vaccine used in this trial is not sufficiently high to suggest that this type of mixed vaccine should be used routinely.

In none of our patients was any reaction noted to the mixed or double vaccinations. Further information is required on the reaction of both negro and white races to combined vaccinations with yellow fever and vaccinia viruses. It is suggested that a careful follow-up of persons vaccinated by the combined Dakar vaccine might produce some valuable information in this respect.

Our thanks are due to Dr George Campbell Young for his very great co-operation, and to Messrs W. A. Whittaker, L. E. Hewitt and D. Santos for technical assistance.

A relatively simple and safe apparatus for the study of airborne infection is described.

Cloud sampling and assessment techniques are described and critically discussed. Evidence is given concerning the satisfactory reproducibility of in vitro and in vivo assays.

Limitations to the usefulness of the apparatus are noted.