Much work has already been done in connection with the differentiation of true diphtheria bacilli from diphthomorphic organisms. Some observers have maintained that all disease producing strains could be distinguished from non-pathogenic types by their sugar reactions. Others have laid much stress on some particular forms of virulence test. More recently serological classification has been attempted in order to help to solve this problem.

With regard to sugar reactions Graham-Smith (1908) found that certain strains produced acid in media containing saccharose, while Hine (1913), on the other hand, maintained that no true diphtheria bacillus could ferment this sugar. Eagleton and Baxter (1922) have shown that diphtheria bacilli, virulent and non-virulent, ferment glucose but do not produce any change in saccharose. Jordan and his collaborators (1922) have found that virulent and non-virulent bacilli fermented glucose and maltose and produced no change in saccharose and that their action on dextrin was variable.

Work on the serological classification of these organisms has also shown some considerable variation. Havens (1920) isolated 206 strains of diphtheria bacilli. These represented cultures from acute cases, release cultures and cultures from healthy carriers. He found that these 206 strains could be divided into two serological groups, one group containing 169 strains, the other 37. There was no evidence of any cross agglutination. Durand (1920) isolated 255 strains of typical diptheria bacilli and after excluding those strains which would not form stable suspensions divided them into five groups, A. B. C. D. and E. containing 16, 8, 25, 61 and 40 strains respectively. Eighty-seven strains were found to be inagglutinable by these five sera but nevertheless could absorb agglutinins from such sera. His final figures for the five groups of bacilli are 18, 8, 31, 76 and 51. Durand and Guerin (1921) in a further paper showed that small isolated outbreaks were always due to the same type of bacillus. They also found that healthy carriers gave the same type of bacillus as was found in the cases arising from these sources of infection. Bell (1922) isolated 133 strains of B. diphtheriae indifferently collected from cases and carriers. He was able to agglutinate 80 per cent. of these strains by three monovalent sera. Group I contained 17 strains, Group II 8 strains, Group III 80 strains.

In view of the results which have just been described it has appeared important to investigate further a considerable number of strains as encountered in routine work, not only as regards the serological characteristics of the organisms but also in regard to their morphology, their biochemical reactions, and their virulence for guinea-pigs.

In a paper on “The Spread of Bacterial Infection” (Journal of Hygiene, xix. No. 4, March 22, 1921) Dr W. W. C. Topley has shown the results of many months' observation of the progress of an epidemic in a population of mice. The author thus describes the experiment: “This experiment lasted from May 21, 1919, until June 11, 1920.…The epidemic started about the middle of October, 1919, and continued in the manner to be described until the conclusion of the experiment.…The period May 21 to September 17 is omitted.…During this time there was no indication of the active spread of infection.…The whole experiment may be divided roughly into four periods. The first extends from May 21 to Sept. 17, and has been referred to already. From Sept. 18 and onwards, only normal mice were added to the cage. From this date until Jan. 5, 1920, the mice were added in such a way as to keep the total number in the cage roughly constant. From Jan. 6 to April 27, 1920, three normal mice were added each day, except on two occasions when none was added. The number of mice in the cage during this period varied from day to day, reaching a maximum immediately before each considerable wave of mortality, and falling to a minimum just before its cessation. During the final period from April 28 to June 11, 1920, two normal mice were added daily instead of three.”

1. Normal rabbits injected intravenously with diphtheria antitoxin obtained from a horse vary considerably both in regard to the duration of Phase B and the duration and the intensity of Phase C.

2. Phase C, the phase of accelerated loss due to formation of precipitin appears sooner and is more pronounced in older rabbits.

3. The early appearance and more rapid formation of antibody (precipitin) in certain rabbits is indicative of naturally acquired immunity.

4. The highest concentration in the blood of rabbits injected subcutaneously with antitoxic horse serum is seen after 2 to 3 days, when it reaches approximately the same value as that of rabbits injected intravenously, and represents only one-third of the total antitoxin injected. Two hours after subcutaneous injection the antitoxic content is only one two-hundredth of that of a rabbit injected intravenously.

1. The course of disappearance of passive immunity in rabbits injected with diphtheria antitoxin obtained from goats, men, guinea-pigs and cows, consists of the same three phases that follow the injection of horse serum.

2. The rabbits examined were more responsive to goat, human and guinea-pig serum than to horse and cow serum.

3. The course of disappearance of passive immunity in rabbits, horses and guinea-pigs injected with homologous antitoxin, consists of Phases A and B only, and Phase B is far slower than when heterologous serum is injected into rabbits.

4. Sheep and goats eliminate antitoxin obtained from a horse at a very slow rate, and Phase C is hardly detectable.

5. Natural immunity of horses to diphtheria toxin is gradually acquired by a number of increasing responses to external stimuli.

The experimental work on which the paper is based was undertaken with the object of investigating the factors which influence the growth of bacteria in various culture media. The B. influenzae was selected as the best organism because of the comparative difficulty of obtaining growth on ordinary media, and the feebleness of the growth when obtained. The importance of food accessory substances, or “Vitamins” for the proper growth and nourishment of animals is generally recognised, and the influence of the same on the nutrition of bacteria has been occupying the attention of various investigators during the last few years.

There are so many unexplained problems in connection with food poisoning that it is still of importance to study in detail and with care every outbreak. I have perused the records of several hundred food poisoning outbreaks, published and unpublished, and the prevailing impression left is one of regret that so many opportunities of clearing up obscurities and elucidating channels of infection have been lost by omission to push investigations home. The outbreak recorded here is in pleasant contrast and worthy of detailed study. It is noteworthy for two features. The bacillus isolated was not only isolated from the incriminated food but it was recovered from water from a tank in the slaughter-house, a rather unusual feature. This suggests a method of infection which has not hitherto been reported.

Conditions in institutions for children or defectives may become peculiarly favourable to the propagation of hookworm disease. The subject of the present study is a long-standing infection with Ancylostoma duodenale in the Hospital for the Insane at Goodna, Queensland, the largest of the three hospitals for the insane maintained by the State Government. The investigation was of special interest because the institution is in a region in which the general population is free from hookworm disease, owing to low rainfall, and because the infection was with A. duodenale, while the predominating hookworm infection of Queensland is with Necator americanus.

In a preceding paper1, details were given of an investigation which demonstrated that an epidemic of milk-borne enteritis occurring in Aberdeen in 1919 was due to infection of milk with dysentery bacilli of the Flexner type. It was further stated that the epidemiological and clinical features of antecedent epidemics of milk-borne diarrhoea were such as to suggest that the former epidemics were probably also dysenteric infections, although bacteriological investigation had failed to identify the causative organisms.

The presence of antibodies in the body fluids of the offspring of immunised parents has long been studied and variously explained. In 1912, Famulener2 published a clear summary of the work upon the subject up to that date. Renewed interest in the problem has been aroused by the work of the Department of Animal Pathology of the Rockefeller Institute for Medical Research, where the transmission of immunity to the Bacillus abortus is being studied. It has been shown by Little and Orcutt3 that agglutinins for this bacillus are transferred exclusively by the colostrum, since a calf from which colostrum is withheld does not show this agglutinin in its serum, although a high titre of agglutinin may be demonstrable in both serum and colostrum of the dam. Later Orcutt and Howe4, using a method of separating the protein fractions of serum devised by the latter5, 6, showed that the appearance of agglutinins in the serum of a calf fed with colostrum of high agglutinin titre was closely related to the appearance of euglobulin, which is absent from the blood of the newly born calf. Further, Smith and Little7 showed that this absorption of agglutinin from the digestive tract also occurred when the calf was fed with serum rich in that antibody. The association of agglutinin with the globulin fraction of serum, and in particular with the euglobulin fraction, has been demonstrated by Orcutt and Howe4, who found that “the agglutinins are associated with the protein fractions which are precipitated up to and including 16·4 per cent. of sodium sulphate,” and that “a large proportion of the agglutinins is associated with the fraction which is precipitated by 14·2 per cent. of sodium sulphate, euglobulin.” Again, although a milk might show a high titre of agglutinin, with no corresponding increase in the proportion of its globulin, yet the agglutinin was removed from the milk by a concentration of sodium sulphate which removed the globulin fraction of the protein.

1. Evidence is given that there exists a multiplicity of serological groups of B. diphtheriae.

2. Ten of these groups have been studied in this communication.

3. Into these ten groups could be placed all except 16 of 348 strains of virulent B. diphtheriae investigated.

4. A description of the technical difficulties encountered is included in the hope that this may prove of service to other workers.

Some evidence on the antigenic structures of different strains is advanced.

5. The application of this method of investigation to the problems of epidemiology is briefly discussed, and evidence of its value brought forward.