Bacteriologists are not in a position to disregard the limitations recognised by the physiologists. Precise data which are not available from experiments are far from sufficient to explain the properties of living matter; they must be supplemented by vaguer conceptions about the functions of a cell, its internal organisation and its susceptibility to stimulative or inhibitory agencies—conceptions which cannot yet be translated into recognised chemical substances and physical properties. One must therefore start by recognising that bacterial virulence, being dependent on vital processes, cannot be fully explained in chemico-physical terms.

In most European countries the geographical distribution of simple goiter is known from the number of men who are rejected on this account for military service(1,2). There are no similar figures for Britain but what is known about its probable incidence has been discussed by Hirsch(2) and by Berry(3) and in a more recent paper, where it was also shown that there was a geographical distribution of the deaths from “exophthalmic goiter” as recorded in the Registrar General's reports(5). There were still many gaps in our knowledge, especially as regards simple goiter, but there appeared to be some tendency for deaths from “exophthalmic goiter” to be more frequent in areas where goiter was endemic, thought this was contrary to the evidence of most parts of the world where goiter was seriously endemic.

A series of 222 strains has been analysed serologically and 156 have been found to react with one or other of four sera prepared with haemolytic streptococci of scarlatinal origin.

Among the 156 strains the four chief types were distributed in the following proportions: 10 of Type I, 57 of Type II, 45 of Type III and 44 of Type IV.

The remaining 66 strains of the series have not yet been classified; though they certainly include individualistic strains, there may be other types to be defined by further analysis.

The four chief types are well defined and can generally be identified by direct agglutination. Cross-agglutination may cause confusion when suspensions are unstable, in which case resort must be made to agglutinin absorption. In addition, certain Type I strains may also agglutinate with Type III serum.

Analysis of clinical data in a series of 100 cases shows clearly that a fairly severe form of scarlatina with greater tendency to complications is often caused by infection with Type II scarlatinal streptococcus; Type II has so far been found only in association with scarlet fever.

There were nine instances where two or more members of the same family developed scarlatina; in six instances the same type was found in each patient and in three each patient yielded an unclassified strain. In one instance the strains from two members of the same family were different.

Haemolytic streptococcus colonies on fresh blood agar present three chief varieties; there are minor differences between these dependent on the amount of moisture in the medium.

There is a certain correlation between variety of colony and serological type. Type III colonies retain a soft consistency after two days' incubation, while the other types generally become tough and can be lifted from the medium as a whole.

Rough and smooth colonies with differential characters analogous to those occurring in pneumococcus cultures have not been identified. An interesting change in appearance is the production of a clear dome-shaped watery or slightly mucinous colony (3rd variety) in a culture which ordinarily forms either more or less opaque or soft coherent colonies.

There is evidence that the proportions of the four main types are likely to vary in different localities and in different outbreaks of scarlet fever in the same locality.

The importance of a reliable classification of haemolytic streptococci is discussed.

An outbreak of food-poisoning is described in which the B. morbificans bovis (Basenau) was the causal agent. The position of this type in the Salmonella group as classified by Bruce White has been confirmed by analysis of the antigenic properties and agglutination reactions (including absorption of agglutinin) of the strains isolated in the outbreak.

1. The development and use of a medium which has selective properties for the growth of B. typhosus and B. proteus is described.

2. The principle of the method rests (1) on the positive property of the B. typhosus of being able to reduce a sulphite to a sulphide in the presence of glucose, (2) on the inhibitory action on the growth of B. coli of a bismuth sulphite in the presence of a certain excess of sodium sulphite.

3. The media finally developed are made in the following way:

A. To 100 c.c. of a melted 3 per cent. nutrient agar are added 5 c.c. of a 20 per cent. solution of glucose, 10 c.c. of a 20 per cent. solution of sodium sulphite (anhydrous), 5 c.c. of a standard bismuth solution. After boiling for two minutes an addition is made of 1 grm. of exsiccated sodium phosphate and 1 c.c. of an 8 per cent. solution of ferrous sulphate crystals.

Medium B is the same as above with the addition of 0·5 c.c. of a 1 per cent. watery solution of brilliant green. The standard liquor bismuthi is prepared by mixing 60 grm. bismuth citrate with 50 c.c. of distilled water and then with 20 c.c. liq. ammonii sp. gr. 0·880 and finally making the volume up to 500 c.c. with distilled water.

4. On these media the B. typhosus grows readily and forms flat blackdry surface colonies. B. proteus grows on the medium in a non-spreading fashion but does not form black colonies. B. coli either fails to grow or after a period of inhibition forms brown sticky raised colonies.

5. Bismuth media were used in the examination of 31 enteric stools and in 30 instances the infecting microorganism was successfully isolated. Single examinations only were made and the material was usually 24 to 48 hours old at the time of examination.

6. Emulsions of enteric stools which as shown by the usual media contained only a dozen or so of typhoid bacilli were found by our bismuth media actually to contain several thousand.

7. The isolation from a case of typhoid fever of a proteus X 19 strain is recored.

8. As regards its growth on bismuth sulphite media B. paratyphosus B behaves more like a reducing B. coli than a B. typhosus culture. For the isolation of B. paratyphosus B a lactose bile salt brilliant green medium is described.

1. There is recorded for the first time an outbreak of milk-borne bacillary dysentery which has been proved to be due to B. dysenteriae Sonne.

2. In this milk-borne outbreak of Sonne dysentery there were over 150 cases and no deaths, the absence of mortality being in marked contrast to the high mortality rate which accompanies milk-borne outbreaks of Flexner dysentery.

3. Several milk-borne outbreaks of enteritis simulating dysenteric infections have been recorded in which the evidence has pointed to a living bacillus of unrecognised type as the causal agent. It is obvious that in future B. dysenteriae Sonne will require to be excluded from the group of the unrecognised viruses.

An experiment recently described in this Journal (Greenwood, Newbold, Topley and Wilson, 1926) dealt with the after–life of mice that had been previously exposed to epidemic pasteurellosis, and led to the following deductions:

“We think we have proved—

(1) That the survivors of herd-exposure to epidemic pasteurellosis are more resistant to subsequent exposure than healthy animals not previously exposed to risk.

(2) That this superiority is significantly correlated both with length of previous exposure independently of its severity, and with severity of previous exposure apart from its duration.

We think it is probable that—

(3) The severity of the prior exposure as measured by the average deathrate during the period of exposure, is less important than the length of exposure.

(4) The advantage of exposure at first increases with its duration and then decreases, so that mice who have been exposed for a moderate time are more resistant to subsequent exposure than mice who have been exposed for a very short or a very long time.”

The results of an independent experiment, similar in kind, with the difference that the after-life and previous exposure took place in a single cage, and covering 1867 mice, support(1), (3) and (4) of the above deductions, also (2) so far as it relates to length of previous exposure, but the severity of previous exposure apart from its duration shows a negative but insignificant association with improved expectation of after-life. These results are in agreement with the suggestion inspired by the experiment previously described. “That these facts are difficult to interpret in terms of pure selection and more easily reconciled with a process of active immunization during the primary exposure, but that the nature of this process will remain obscure until we have more experimental data at our disposal.”

1. The three organisms B. pyocyaneus, B. fluorescens-liquefaciens and non-liquefaciens form a closely related group of organisms both biologically and also by reason of their behaviour towards disinfectants.

2. Bacteriological observations support the view that germicides can naturally be divided into two classes, viz. “chemical” and “physico-chemical,” their actions on micro-organisms depending respectively on

(1) Chemical reactivity with protoplasmic constituents.

(2) Denaturation or precipitation of the cell proteins.

3. B. fluorescens non-liquefaciens and related organisms are selectively attacked by “physico-chemical” germicides, e.g. phenol, whilst B. coli is less sensitive to their action, but is particularly susceptible to the “chemical” disinfectants, e.g. quinones.

The B. fluor. non-liq. group is also more readily affected by hot water than B. coli.

4. Substitution of various groups into the benzene nucleus in the case of aromatic disinfectants affects the germicidal power on B. coli and B. fluor. non-liq. unequally, with the result that the selective action and differentiation may be obscured or even reversed in the case of the substituted derivatives.

5. B. fluor. non-liq. is more sensitive than B. coli to the action of reducing agents, and this factor also affects the degree and nature of selective action, sometimes resulting in the classification of “chemical” germicides in the “physico-chemical” group.

6. Selective action is not affected by changes in experimental conditions, such as temperature and period of disinfection, nature of culture medium, number of bacteria present, or physical state of the disinfectant. The susceptibility of B. fluor. non-liq. to the action of hot water, and such germicides as phenol and alcohol, is thus probably associated in some way with its internal colloid state.

7. The analogy between the action of hot water and that of the alcohols and phenols is borne out by preliminary experiments with the thermophile bacteria (heat resistant). These organisms are somewhat less sensitive to phenol that B. coli, and more susceptible to p-nitrosodimethylaniline.

Since disinfectants containing hypochlorites have been used for the sterilisation of dairy utensils, and it has been claimed that such disinfectants are more efficient than steam sterilisation, it seemed advisable to study the question in the interests of the Milk Industry. The efficiency of three proprietary disinfectants (A, B and C) has, therefore, been investigated.

The following investigation into the effect of pasteurisation on the tubercle bacillus was commenced in 1923, in order to ascertain whether the temperatures used in commercial pasteurisation are at all times sufficient to destroy the virulence of the tubercle bacillus in naturally infected tuberculous milk, i.e. milk as secreted by a cow suffering from tuberculosis of the udder.

(1) Non-specific agglutination as tested by thermo-agglutination, peptone agglutination, lactic acid agglutination and Michaelis' acid agglutination falied in the primary object of differentiating between Br. melitensis and Br. abortus.

(2) Thermo-agglutination was only observed in strains of Br. paramelitensis.

(3) With peptone agglutination no strain gave a positive result.

(4) The results of lactic acid agglutination and Michaelis' acid agglutination ran practically parallel. No distinction could be observed between Br. melitensis and Br. abortus. Strains of both of remote origin gave slight agglutination with lactic acid while recently isolated strains of Br. abortus were negative. Very definite agglutination, however, was observed by both methods with paramelitensis and paraabortus strains, but no differentiation was possible between such strains.

1. A case of the septicaemic form of melioidosis is recorded from Cochin China.

2. Certain new biological characters of the causative agent, B. whitmori, are described.

3. An animal origin of the infection in this case is considered probable.

The biological characters, and pathological effects in animal inoculation, of a strain of B. whitmori from Saigon are described. It is the most virulent strain of the organism from a human source that has so far been encountered. The case from which the organism was recovered was an example of the septicaemic from of melioidosis.

(a) The slow lactose fermenting bacilli which we have investigated cause an acute febrile illness in man. The general symptoms may be so severe that the diagnosis of typhoid and paratyphoid fever has been made. Such cases are invariably due to an acute infection of the urinary tract, more specially acute pyelitis or pyelo-nephritis, but the general symptoms may be so severe at the onset of the illness that the urinary manifestations are masked.

(b) Three fatal cases have occurred.

(c) In cases of acute diarrhoea, with or without blood and mucus, these bacilli may be present in pure culture in the faeces.

(d) Chronic urinary infections due to these bacilli are uncommon.

(e) The bacillus of so-called columbensis fever is allied to these bacilli.

The analysis of an epidemic of influenza in the New Zealand division of the Royal Navy confirms the following hypotheses which were suggested by previous studies of the same and other diseases.

(1) The type of influenza in different places, especially if they are more or less isolated areas, such as islands or ships, tends to assume special characters as regards infectivity, virulence, and clinical type1. In this respect the influenza virus, or viruses, are only following a well-known biological “law,” namely, that isolation tends to encourage the evolution of new varieties and species. A classical example of this “law” is found in Darwin's reference to the fact that each island of the Galapagos Archipelago has its own special flora and fauna—similar to but yet different from that of the other islands.

(2) One wave of influenza, even if of a different type, will confer considerable immunity to a subsequent wave not only on the victims attacked, but also on those who apparently escaped infection during the first wave.

(3) The brunt of the morbidity of influenza (and other infections) is borne by recruits and junior ratings, not because they are, on the whole, younger, but because they have had less time than the senior men to adapt themselves, by means of auto-vaccination, to the bacterial environment of crowded ships or barracks.

The effect of bacterial adaptation to environment is often overlooked. Hence other factors, such as physical fitness, are often given more credit than they deserve as preservers of the public health.

(4) Senior ratings mixed with recruits suffer a higher morbidity from infectious disease than they do under similar circumstances in the absence of recruits. This statement, if true, should have a most important practical bearing on military medical administration. The phenomenon is probably due to increased velocities of infection (due to the higher infectibility of the recruits) breaking down the herd immunity of the senior men.

(5) In this epidemic New Zealanders suffered more from influenza than the British sailors not, as usually supposed, because of the mere fact of being born in New Zealand, but because all the recruits and the bulk of the junior ratings were New Zealanders.

1. Examination of nine type strains of Brucella, collected from various sources, by agglutination and the absorption of agglutinin test revealed distinct serological difference between Br. abortus and Br. melitensis. Differentiation could only be made by the agglutinin-absorption test, both organisms agglutinating equally with abortus or melitensis immune sera.

2. By agglutination alone differentiation could be made between a group consisting of Br. melitensis and Br. abortus strains on the one hand and a group which included one strain labelled Br. melitensis, one labelled Br. abortus, and Br. paramelitensis.

3. The agglutinin-absorption test showed that this Br. paramelitensis group comprised two strains of closely allied paramelitensis strains, and that the Br. abortus strain could be regarded as a Br. para-abortus strain.

4. Serological investigation of eight strains of Brucella isolated from patients suffering from undulant fever in Southern Rhodesia showed that six were serologically identical with type Br. abortus and two identical with what is regarded as a Br. para-abortus strain.

5. It is suggested that the melitensis-abortus group represent the “S” normal types, whereas the paramelitensis-para-abortus group represent the “R” mutant types of organisms.

A race-horse, which had been imported into the Malay States from Australia, got out of condition and occasionally coughed up thick pus. Five specimens of pus were brought to the laboratory between February and July, 1925, and on each occasion B. whitmori was isolated from it. The blood of the horse agglutinated the standard strain of B. whitmori in high dilutions. The morpho-logical characters, and the biological reactions of the organism isolated from the horse were the same as those of the type strain (Ragaviah) of B. whitmori; and cross-agglutination and cross-absorption tests showed that they were identical. The strain was highly pathogenic for guinea-pigs and rats. The horse was killed in September, 1926, eighteen months after B. whitmori was first isolated from the nasal pus. No active lesions were found at autopsy and B. whitmori was not isolated from the organs.

In a preliminary paper (1926) I have shown that most strains of haemolytic streptococci obtained from scarlet fever may be placed in two very definite serological groups. The strains from 119 cases were found to belong to Type I, from 57 cases to Type II, while from 34 cases the strains were un-identified. That the scarlatinal streptococci do not form one serological group has been confirmed by Griffith (1926) and by James (1926). Further, the sera and type cultures used in this investigation have been interchanged with those of Dr Griffith, and his Types I and II strains have been found to conform with the main serological types found in Aberdeen.

In the year 1911 the heading “Status Lymphaticus” was first used in the publications of the Registrar-General of England and Wales and, in that year, 121 deaths were assigned primarily to this heading; Status Lymphaticus was also mentioned in the certificates relating to 25 of the deaths which occurred under anaesthesia. The heading was retained in the published tabulations as first introduced until 1921, when certifications of Status Lymphaticus were classified under the heading of “Diseases of the Thymus,” of which statistical group they have, so far, formed the whole. Between 1911 and 1924 (inclusive) the annual deaths—inclusive of those associated with anaesthesia—have fluctuated between 145 and 273. In the last four years there has been little variation, 209 in 1921, 204 in 1922, 200 in 1923 and 199 in 1924. Most of the deaths are in the age group 0–5 and a considerable majority are of males. Thus in the last two available years, 1923 and 1924, of 399 deaths 289 were at ages 0–5 and 258 were of males (respectively 72 and 65 per cent.). If the 85 deaths in which anaesthesia was an associated condition are excluded, 81 per cent. are of children aged 0–5 and 64 per cent. of males (Table I).

(1) Strains of B. pestis, kept in culture for 7 and 9 months, showed no loss of virulence.

(2) Virulent strains, when subjected to passage through an immunised Cricetomys gambianus (African pouched rat), suffer distinct loss of virulence for this species, and virulence is not regained in culture or on repeated passage through susceptible C. gambianus. The so-called avirulent plague bacillus, obtained in this way, appears to be a rough variant. It differs in micro-scopical appearance from the ordinary virulent forms, tending to the appearance from the ordinary virulent forms, tending to the appearance of involution forms, which develop in salt media.

(3) As regards immunising properties:

(a) Vaccine prepared from virulent strain gives considerable protection against infection with virulent strain.

(b) Vaccine prepared from virulent strain gives no protection against infection with avirulent strain.

(c) Vaccine prepared from avirulent strain gives little or no protection against infection with virulent strain.

(4) Animals, which recover from infection with an avirulent strain, are immune to subsequent infection with virulent strain.