The manuscript notes from which the following Account of the Journal of Hygiene has been prepared for the press were left unfinished by Dr Graham-Smith at the time of his death on 30 August 1950. He intended that the Account should be published to mark the fiftieth year of the Journal's existence.

G. S. Graham-Smith was born on 25 September 1875, the son of a colonel in the Indian Army. He expressed a wish, during his life-time and in his Will, that no obituary notices should be published about him, because so few of those with whom he was intimately associated during the most active period of his working life now remain. It has, nevertheless, seemed proper to submit the following account of him, because many friends and students and some contributors or subscribers to the Journal who knew him only as editor, will wish to see such details of his life and work recorded.

Small enclosed populations of the common brown rat (ten to twenty-six rats), each with access to one ton of sacked wheat for 12–28 weeks, caused a loss in weight of 4·4% of the wheat. 70·4% of the wheat was fouled and had to be cleaned before use. The main monetary loss was due to damage to sacks. Total monetary loss was 18·23% of the original value of the wheat and sacks.

A transformation devised by Mather is applied to give a new scheme for computing the maximum likelihood estimate of a bacterial density from the evidence of a dilution series. Tables are provided to expedite the application of the method; with their aid, the calculations take a from similar to, but much simpler than, those for probit analysis of quantal responses.

Maximum likelihood estimation is compared with the method proposed by Fisher. The latter has the advantage of extreme simplicity, at least for series to which existing tables can be applied, and the loss of information involved in its use may often be compensated by the saving of time in calculation. An experimenter who has fairly reliable prior indications of an approximate value for the density, however, ought to concentrate his attention on dilutions that he believes will contain between 4 and 1/4 organisms per sample; he must not apply Fisher's method to his results, but the maximum likehood estimate will be so much more precise than any estimate from an experimental design using more extreme dilutions as to repay the small additional labour in computation.

1. The injection of mice with botulinum spores, freed from all traces of superficial toxin, can cause death but only if a very large number is given. This number was found in the case of a highly toxic strain to be of the order of 20 million spores given intramuscularly. This dose was fatal in the absence of spore germination.

2. In human botulism it is concluded that the spores of Clostridium botulinum persent in spoiled foodstuffs are not of importance compared with the toxin produced by the initial growth of the contaminating organisms.

3. Toxin could be demonstrated in the cytoplasm of botulinum spores when these were disintegreted by mechanical means. This is of theoretical interest as the toxin within the spores is more heat-stable than the extracellular form.

4. A local anaerobic myositis could be produced in mice by the intramuscular injection of spores plus 2·5% CaCl2. The number of viable spores in the LD50 was of the order of 50,000.

5. Calcium chloride solution in the commonly used strength of 2·5% was shown regularly to cause a drop in the number of botulinum spores in a given suspersion of some 40% and a parallel loss of viability of about 60%.

Actinomyces bovis passes through a complete life cycle. The spore germinates to produce a haploid generation, from which a diploid generation arises by conjugation of two specialized haploid cells. A haploid spore is produced from the diploid generation by reduction division.

1. Evidence is submitted that an anomalous rabbit cell lysin reported by Minett (1936), α2-toxin and δ-haemolysin are identical. The last name is preferred.

2. Precautions are necessary to avoid interference by δ-haemolysin in titrations of α-toxin and α-antitoxin.

3. The existence of γ-haemolysin is confirmed; its wide distribution and possible confusion with α-toxin are noted. The name γ-toxin is thought more suitable than γ-hameolysin.

4. Multiple standard antisera are desirable for the detection of anomalous toxins. At least one should be of refined globulin.

5. Dilution and manipulation inactivate staphylococcal toxins and their effect needs consideiration in quantitative work. Infusion broth increases the stability of diluted α- and β-toxins.

A study was made of the reaction of the white blood corpuscles in bovines at parturition.

Graphs made from the differential total counts of the white blood corpuscles from shortly before calving and at intervals through the period of parturition and for 24–48 hr. after calving show a definite absolute lymphopoenia with a rise in the neturophils.

The type of graph produced resembles that found in bovines when cortisone is injected.

The peak of the white blood cell reaction usually takes place about 5–9 hr. after the delivery of the calf. This is also the period of the peak reaction in the cellular relations after injection of cortisone.

The observations recorded seem to afford positive evidence that C 11 hydroxy steroids (cortisone) are poured out in the last stage of parturition in the normal bovine. The desensitization of the skin in animals which were sensitized to Trichomonas foetus an which was found at this time is considered to be connected with the action of the hormone.

The Cambridge University Health Service started work in October 1948. So far it has been restricted to preventive measures and has been confined to new entrants to the University on an entirely voluntary basis. The Service provides a clinical examination and a chest X-ray. A general survey of the scope of the Service and the results obtained have been given in the annual reports of the Senior Health Service Officer, Sir Alan Rook (1950, 1951). In addition to the original clinical records a Hollerith punched-card record system has been used since the inception of the Service to facilitate detailed investigations. In Appendix I to the first annual report mentioned above (Rook, 1950) I gave a few preliminary results on various physical measurements for the first 900 men examined. It has now been possible to complete a much more extensive investigation, the results of which are given in the present paper. In some cases this has been based on the completed records for the year 1948–9, while in others the records for the two years 1948–50 have been utilized. I have already given a brief summary of the contents of this paper in Appendix I to the second annual report (Rook, 1951).

1. The most virulent strains of Salmonella typhi are those that develop both the Vi and O antigens in maximum quantities.

2. The Vi antigen is not a particularly toxic substance; it is less toxic for the mouse than the O antigen.

3. The main biological function of the Vi antigen is to protect the O antigen against the action of the natural or immune O antibody, thereby protecting the bacterial cell against phagocytosis and the bactericidal action of the serum.

4. The relative virulence for mice of freshly isolated strains of Salmonella typhi reflects their virulence for man.

Intraperitoneal inoculation of the mouse has been employed for many years as a means of assaying the virulence of different strains of Salmonella typhi or of detecting alterations in virulence in the same strain occurring from time to time in the laboratory. Such changes are usually in the nature of loss of virulence resulting from antigenic change. Occasionally, however, there may be increase of virulence, either as a result of mouse passage or simply from a more strict observance of nutritional requirements and improvements in the method of cultivation. Where cultures of typhoid organisms are to be used for vaccine production the detection of these antigenic changes and the associated loss of virulence is obviously of the greatest practical importance.

A bacteriophage is described which attacks almost exclusively strains of Bacillus anthracis, and attacks all the strains of this species which have been tested.

As well as attacking typical strains of this species it attacks atypical variants.

Strains of B. anthracis are not attacked by this phage when they are in a muciod state.

A few strains of B. anthracis are only feebly attacked by this phage. Some experiments, conducted in an attempt to discover the reason for this, are described.

The results are discussed and arguments in favour of regarding B. anthracis as a valid species are put forward.