1. The rigidity of agar gel in poured agar plates can be tested by a comparatively simple and easily repeatable method.

2. Commercial agar from different sources varies in the rigidity of gel it produces at a given concentration.

3. The rigidity of gel produced by any agar at a given concentration is affected by processing and by the mode of production of poured plates and to some extent by the diluent used.

4. The appearance of a spurious S-R change in colonies of Salmonella and Shigella on media with an increasing agar content is a measure of the rigidity of the agar gel used and not directly of the agar concentration of the medium.

5. The appearance of this spurious S-R change induced by rigid agar bears no relation to the agglutinability of the organism.

6. Both increased or decreased agglutinability of cultures of Salmonella and Shigella is associated with the appearance of colonies of an apparently rough morphology at a lower level of agar rigidity than that at which spurious S-R change develops in a normally agglutinable culture.

7. Without some standardization of the rigidity of the gel used in agar plates, erroneous conclusions may be drawn regarding the antigenic state of the organisms under test.

8. Samples of agar of differing gel rigidity can be blended to form a product producing a gel of a calculated 12 sec. value.

I am indebted to the Director-General, Army Medical Services, for permission to publish this article.

The type strains of Bact. coli, described by Bray (1945), Giles & Sangster (1948), Giles et al. (1949), Taylor et al. (1949), Smith (1949) and Rogers (1951b), because of their very intimate association with epidemic infantile gastro-enteritis, have been used as indicator organisms to show that the method of cross-infection of epidemic gastro-enteritis is very like that of the cross-infection of burns.

It has been shown how a cubicle becomes widely contaminated within 18 hr. and that type strains of Bact. coli will remain viable in dust for at least 27 days. Communal articles used throughout a ward are blamed for the inter-cubicle spread in a divided ward.

The intimate association of the α- and β-types of Bact. coli with outbreaks of gastro-enteritis is demonstrated, and figures are given to show how these outbreaks prolong the patient's stay in hospital and waste valuable hospital space.

Evidence of how epidemics of infantile gastro-enteritis can spread from one hospital to at least three other hospitals has been presented. Three such outbreaks are described, in two of which the α-type of Bact. coli, and in one the β-type of Bact. coli, were associated with the cases of enteritis.

It is suggested that in hospitals which admit infants routine examination of the infants' faeces should be made to identify these types of Bact. coli, and that institutional epidemics of gastro-enteritis should be notified to a central bureau together with the cultural findings, so that the pooled knowledge could be redistributed to paediatric centres.

The aetiological relationship between infantile gastro-enteritis and these type strains of Bact. coli is briefly discussed.

The prevalence of the different types of coli–aerogenes bacteria has been determined in various soils collected from widely separated locations in the United Kingdom and in Denmark. Of the soils examined, 27% contained no coli–aerogenes organisms in 15 g. of soil tested. The number of these organisms in the positive soils was very low; only 11% had confirmed counts greater than 100 organisms/g. and 33% less than 1 organism/g.

The most prevalent organism isolated was Bacterium coli type I, whatever the location or vegetation type of the soil; intermediate type I, Bact. aerogenes type I and Bact. coli type II were the only other organisms of numerical importance. The annual application of large amounts of either farmyard or inorganic manures produces no permanently large population of coli-aerogenes organisms in soils.

Presumptive tests in MacConkey broth were unreliable, since only 6 % of the positive tubes from hill and moorland soils and 28 % from arable soils were confirmed as containing coli-aerogenes bacteria. Such discrepancies are considered to be partly due to coliform bacteria with weak lactose-fermenting properties.

It is considered that there is now sufficient experimental evidence to discredit statements that soil is the natural habitat of Bact. aerogenes or intermediate types of coli-aerogenes organism.

The lysis of Bacterium coli suspensions brought about by glycine shows the following characteristics:

(1) There is a latent period of 2 hr., followed by a rapid lysis reaching a maximum in about 8 hr.

(2) The extent of the lysis is independent of the dilution of the bacterial suspension over a wide range.

(3) The extent of the lysis increases with the glycine concentration up to 10M, but is approaching a limit at this concentration.

(4) The lysis is negligible below pH 5 and above pH 10, and shows a maximum rate in the region of pH 6–5–8–5.

(5) The rate of lysis has a very high temperature coefficient (Q10 of the order of 5).

The life cycle of Nocardia starts-with the microcyst, which germinates by budding. The bud develops into a long multinucleate filament, which becomes divided into individual cells by the formation of transverse cell walls. Microcysts are formed in the same manner as in the Eubacteria. Multiplication of the cells takes place by simple fission, by complex vegetative reproduction, and by branching. They do not have a prolonged diploid phase like the Actinomyces.

1. The effect of virulent bovine tubercle bacilli on rabbit tissue in vitro was studied.

2. When normal spleen cultures were infected in vitro the cells actively phago- cyted the bacilli. At first the phagocytes tolerated the organisms, and in the course of their wanderings spread the infection throughout the tissue. Some infected cells even underwent mitotic division.

3. Eventually the infected cells broke down, and the freed bacilli continued to grow in the tissue. Fourteen days after infection, outgrowth had usually ceased and the explants were almost completely replaced by bacilli.

4. Actively growing spleen cultures were unable to suppress even a slight infection with virulent bovine bacilli.

5. No tubercles were formed in the cultures.

6. There were no indications ofan exotoxic action of the bacilli on the cells in tissue culture, an densely infiltrated explants were often surrounded by a large zone ofoutgrowth.

1. Rabbits were infected with virulent bovine tubercle bacilli in three ways: by (1) intravenous, (2) intrasplenal and (3) intrapulmonary injection. In series (1) and (2) part of the spleen, and in series (3) part of the draining lymph glands were sectioned for histological examination, and usually explants for tissue culture were made from the rest of the organ.

2. A histological study of the infected spleens and lymph glands fixed after different periods of infection showed the usual picture of progressive tuberculous infection; acid-fast bacilli were rare and often difficult to find.

3. After a few days' cultivation in vitro, explants of the infected spleens often became densely infiltrated with bacilli.

4. Previous investigators have expressed the view that acid-fast bacilli are so few in tuberculous lesions in vivo, because large numbers become unstainable. If this is true, and if the unstained bacilli remain viable, it would readily account for the extraordinarily rapid bacillary infiltration of tissue cultures of infected organs. Alternatively, it must be assumed that the few scattered acid-fast rods in the original tissue grow at a prodigious rate under the conditions of life in vitro.

5. These alternatives and other theoretical implications of the results are discussed.

Discrete, Feulgen-positive granules, attached to the periphery of bacterial spores, were described by Pietschmann & Rippel (1931), who considered them to be residual cytoplasmic structures; by Stille (1937), who was apparently in doubt as to their nature; and by others, including Robinow (1945), who at first described them as the natural appearance of the spore nucleus. Delaporte (1950) figured the spore nucleus as a less discrete, crescentic body lying against one side of the spore, and in a later study Robinow (1951) accepted this view, and presented evidence to prove that the appearance of the discrete body was an artefact, caused by the hydrolysis with HC1 which was employed in staining. Robinow, in this later study, employed nitric acid instead of hydrochloric acid, and claimed that it gave a truer picture. Both Delaporte and Robinow observed a central, stainable body, which they considered to be cytoplasmic.

A series of thirty-five patients suffering from diphtheria, who had been previously fully inoculated against the disease and from whom toxigenic diphtheria bacilli were isolated, were investigated and compared with seventeen patients who had not been inoculated. There were smaller groups of those patients who had had incomplete courses of immunization and of those who had had previous attacks of the disease.

Two-thirds of the infections in the inoculated were due to Corynebacterium diphtheriae, intermedius type. Intermedius infections were proportionately more numerous in the inoculated than in the uninoculated, whereas gravis infections were less frequent.

Estimations of the antitoxin in the blood were made on admission. The inoculated patients had levels of antitoxin below, or in the neighbourhood of, that believed to confer protection.

In the inoculated group severity did not appear to be influenced by the level of antitoxin in the blood on admission.

Antitoxin in the blood was found to have increased after recovery from the disease in all except one patient; the increase was considerable in those who had been inoculated previously.

From 23·5% of all cases of clinical diphtheria investigated only non-toxigenic diphtheria bacilli (gravis, intermedius and mitis types) were isolated. These cases were considered separately. On admission the majority had in their blood considerable amounts of antitoxin, which did not increase after recovery. No other pathogenic organisms were regularly found. The possibility is discussed that under certain conditions non-toxigenic C. diphtheriae may be able to cause a diphtherialike disease.

An investigation of the survival rate of bacteria after drying in vacuo by the ' spin–freeze' method showed that one of the most important factors was the fluid in which the bacteria were suspended. Broth or nutrient gelatin gave a moderate survival rate, but in various other protein solutions, even though there might be a fair survival immediately after drying, a large fall occurred in the viable count after storage. The addition of glucose to the suspending fluid in concentrations of between 5 and 10% greatly increased the survival rate both immediately and after storage. Some experiments showed that, without added glucose, the longer the drying process was continued the lower was the survival rate, and it is thought that glucose acts by automatically retaining a small amount of water which is necessary for survival. Lactose appears to produce the same effect.

It is shown that the highest death-rate generally occurs during the early stages of drying, the further drop after 24 hr. being small if glucose is used in the suspending fluid. Very young cultures (4½–6 hr.) seem to be far more sensitive to drying than older ones (18–24 hr.). The bacteria that survive drying are not more resistant to subsequent drying.

Most of these results have been obtained with one strain of paracolon bacillus, but a limited number of other experiments show that the same principles hold good for other organisms such as Salmonella typhi-murium, staphylococci, streptococci, Neisseria gonorrhoeae and Vibrio cholerae. In the case of the last two species the percentage survival is much lower than with the others, but survivals up to 1% have been obtained after storage for from 1 to 3 years.

A new method of titrating the infectivity of egg-adapted influenza virus suspensions is described.

For this purpose a plastic tray has been made, drilled with one hundred conical cups. In each cup a small piece of chorio-allantoic membrane from a chick embryo is suspended in a buffered salt solution. To these surviving tissue suspensions virus dilutions are added, and the tray is put in an incubator and rocked mechanically for 60 hr.

In those cups in which the inoculum has been sufficiently large, the membrane becomes infected by the virus which multiplies within the cells and, bursting out, builds up by infection of other cells a high concentration of virus in the suspending fluid. The virus multiplication is detected by removing the pieces of tissue and adding to the cups a drop of washed chicken erythrocytes. The presence of virus in the cup is shown by the pattern of the sedimented erythrocytes, and virus multiplication is inferred because the original seed virus is too dilute to be detectable by haemagglutination.

The method has been compared with in ovo titration. As a means of detecting egg-adapted infective virus, the tissue suspension is between 20 and 80 times less sensitive; as a titration system, however, approximately one cup is equivalent to one egg, so that by using sufficient replicates accurate assays of infective virus can be made.

The technique is readily adaptable to the titration of serum-neutralizing antibodies.

In addition to haemagglutination, the infectivity end-point of the titration can also be estimated by a specific complement fixation test, and it is suggested that, with modifications, the new technique may be applicable to the titration of a large number of other viruses.

1. A series of 110 patients with staphylococcal infections has been investigated. From 43 patients a penicillin-resistant staphylococcus was isolated: in cultures from 38 patients all staphylococcal colonies tested were penicillin resistant and in five both sensitive and resistant strains were isolated. Seven patients yielded a streptomycin-resistant staphylococcus.

2. Of the 84 strains of staphylococci which could be phage typed, 33 belonged to the group of phages 6/7/47 and 32 were of type 52 A. Of the 32 typable penicillin- resistant strains 18 belonged to the group of phages 6/7/47, and 13 were type 52 A.

3. Of the 115 strains isolated 110 were α-lysinogenic.

4. These results are compared with those found a year previously and the significant changes are discussed.

1. The review of the laboratory evidence published since the introduction, ten years ago, of the alcohol-treated typhoid vaccine furnishes additional support for abandoning the old method of making typhoid vaccine. The alcohol-treated vaccine has advantages in regard to both the Vi-antibody response and the degree of systemic reactions produced.

2. The technique of the preparation of alcohol-killed and alcohol-preserved typhoid-paratyphoid vaccine has remained as originally described. The methods of selecting the vaccine strains and routinely examining the cultures have also remained unchanged.

3. The necessity of testing typhoid vaccines not only by active-immunity tests in mice but also for their antibody-stimulating properties is again stresssed. These tests consist of immunization of rabbits, estimation of their Vi- and O-agglutinin titres and passive-immunity tests in mice.

4. The most important sources of error in mouse-protection tests are:

(a) the use of a test culture of less than the maximum degree of mouse-virulence;

(b) the use of the mucin technique;

(c) immunization by intraperitoneal injections when the challenge dose also is given by this route.

Much of the experimental work on typhoid vaccine has been invalidated by these three pitfalls.

5. It is suggested that the mucin technique be abandoned in the assay of typhoid vaccine.

6. A source of error in the rabbit test is the presence of pre-formed Vi agglutinins. These are found in rabbits harbouring coliform organisms which possess the typhoid Vi antigen. Such rabbits do not respond to injections of typhoid Viantigen.

7. Standardization of the potency of typhoid vaccine has now become possible. Either an alcohol-preserved or a dried vaccine can serve as ‘standard vaccine’, to be used in combination with the ‘Provisional Standard Anti-typhoid Serum’. Both of these vaccines remain stable for a number of years.

8. Suggestions are briefly outlined for:

(a) the routine examination in the course of preparation of the vaccine;

(b) the official control of the potency of typhoid vaccine.

9.The paratyphoid components of T.A.B.C. vaccine can be standardized in an analogous manner.

1. Alcohol-killed and alcohol-preserved typhoid vaccine retains its power of stimulating Vi- and O-antibody formation unimpaired during storage at 1–2° C. for at least 10 years.

2. The Vi antibody induced by immunization with the 10-year-old vaccine possesses full ‘functional efficacy’ in passive-immunity tests in mice.

3. The expiry date of this type of vaccine can be extended to a period of 10 years, provided the vaccine is kept in cold storage.

4. Alcoholized vaccine can serve as ‘standard vaccine’ in any future attempt at standardization of the potency of typhoid vaccine.

1. A double enteric infection is defined as the simultaneous infection of an individual or group of individuals with two organisms of the enteric group. The literature of the previously recorded cases and epidemics, in which the diagnosis of double enteric infection has been established by cultural methods, is reviewed.

2. An account is given of a double enteric outbreak of seventy–six cases which occurred amongst British troops and police at Acre, Palestine, in 1948. The infection is thought to have been due to contamination of the water supply with sewage during the civil disturbances. The diagnosis was established bacteriologi–cally in seventy-four cases (93.3%);Salm. typhi was isolated in forty–three cases, Salm. paratyphi B in three, and both Salm. typhi and Salm. paratyphi B from twenty–eight cases. There were three fatal cases in the epidemic (mortality rate = 3·94%).

3. The morbidity rate from enteric fever among the infantry unit was 60%, and among the Palestine police, 17%. The possible reasons for this difference is discussed, and the conclusion drawn that it was probably due to the Palestine policemen being older men, with longer overseas service and more 'seasoned' to life in subtropical conditions.

4. Thirteen cases were treated with polymyxin B (15 mg. 4-hourly for 4-day periods); with this small dose no beneficial therapeutic effects were observed, and ten cases showed evidence of renal damage while under treatment.

5. The bacteriology, epidemiology and clinical aspects of double enteric infections are discussed. The diagnosis of a double enteric infection may be established with the greatest certainty by blood culture. Such infections are usually water-borne or milk-borne, and tend to occur when there has been a severe breach of hygiene, e.g. in the contamination of a water supply by sewage. The claim that the prognosis in instances of double enteric infection is worse than with single infections is not supported by the experiences at Acre where the three fatal cases occurred in cases infected with Salm. typhi alone.

1. The results of 192 blood cultures in an epidemic of enteric fever are analysed. The epidemic was caused by a double enteric infection with Salm. typhi (Vi-phage type T) and Salm. paratyphi B (Vi-phage type ‘Dundee’).

2. All of sixty –five cultures between 102 and 104∘ F. were positive, and a significant number of positive cultures at lower body temperatures were obtained. Three out of four cultures, taken when the temperature was subnormal, were positive. In seventy–one out of seventy–six cases (93·4%) a positive blood culture was obtained.

3. The highest percentage of positive isolations was obtained in the second week of disease. In the severe cases the highest percentage occurred in the third week. The reasons for this, and the findings of previous studies, are discussed.

4. No difficulty was encountered in obtaining positive blood cultures in previously inoculated individuals.

5. The average period of incubation of the cultures, required to produce a positive result, was 4–5 days, and 17% did not become positive until the ninth to eleventh day of incubation.

6. The discovery of ehloramphenicol has emphasized the need for the early diagnosis of enteric fever. This can most readily be achieved by the method of blood culture.

1. An epidemic of typhoid fever involving 135 persons in a semi-closed community of. 742 persons is described.

2. There were nineteen symptomless excreters (14·1%).

3. Among the 116 clinical cases there were seven deaths, giving a case fatality of 6.03%. If the case fatality were calculated on the total persons involved (including symptomless excreters), the figure would be 5·2%.

4. The origin of the outbreak was not discovered, but there was strong cricumstantial evidence in favour of a milk-borne infection.

1. An account is given of the finding of 19 symptomless excreters of Salm. typhi and of 15 incubating cases (‘precocious carriers’) in the course of repeated examination of the faeces of 435 staff and patients at the hospital.

2. In a comparison of the frequency of isolation of Salm. typhi from the faeces it was found that the the selenite F enrichment method was more efficient than enrichement in Kauffmann 's tetrathionate broth, and that both methods were superior to direct plating on desoxycholate citrate agar.

3. Examination of sera from cases of typhoid fever showed that, in some cases, H- and O-agglutinin titres which would normally be regarded as ‘diagnostic’ may be reached well within the first week of illness.

4. Measures are described which were taken to exclude connexion between this outbreak and (a) a chronic carrier found during investigation of the hospital milk supply, and (b) the incidence of typhoid fever at another hospital.

5. A case of typhoid fever at Shrewsbury is described which was probably infected by an abortive missed case. Attempts to make a retrospective bacteriological or serological diagnosis of the latter were unsuccessful.

6. An account is given of the use of the sewage-pad method in an attempt to trace the source of infection in the case of a sewer worker who developed typhoid fever 14 months after the main epidemic and was probably infected by an intermittent carrier at the hospital.

1. The low Vi-antigen content and low mouse virulence of freshly isolated cultures of the Oswestry strain of Salm. typhi reflected its low virulence for man. This is additional evidence of the association between the degree of virulence of Salm. typhi and its Vi-antigen content.

2. The prediction that the epidemic caused by this typhoid strain of low mouse virulence would resemble a paratyphoid-B outbreak more closely than an average typhoid outbreak proved to be correct.

3. Sera from convalescent patients in this typhoid outbreak showed significant Vi-agglutinin titres in spite of the low Vi-antigen content of freshly isolated cultures.

4. The nutritional requirements of the Oswestry strain were unusual, and its sensitivity to the specific Vi-typing phages was variable. This caused considerable difficulty in the typing of the cultures.

5. Additional tests of identity had to be employed in order to distinguish the epidemic strain from other strains of the same Vi-phage type isolated in the vicinity. Colony size, colonial morphology, biochemical reactions and sensitivity to anti-O bacteriophage served as useful ancillary criteria.