For the measurement of the temperature of exposed skin surfaces, we have reached the conclusion that the most accurate instrument is a radiation thermopile suitably screened from the effects of wind. As the thermopile does not actually touch the skin it does not interfere with the cutaneous circulation; neither does it hinder the heat-loss from the skin surface. Readings are simply and rapidly made; with the Moll thermopile and the Cambridge unipivot galvanometer used by us a reading can be taken in 4–6 sec. The readings are not influenced by wind if the thermopile is screened, and as the temperatures are measured as differences from that of a blackened surface of known temperature, change of air temperature does not upset the results. The apparatus is easily portable.

By rolling a mercury thermometer over a small area of skin the temperature can be estimated with but very slight error. This method is a reliable one for use where other apparatus is not available; but, it has the disadvantage that each observation takes from 1 to 1½ min. Accurate estimates cannot be obtained by this method in the special cases when the skin temperature is subject to rapid variation.

Under steady conditions accurate estimates can also be made by means of a clinical thermometer protected by a cork wedge, but each observation takes from 4 to 5 min.

Of the types of thermo-junction tested by us the simple Lewis type of junction attached and protected by a strip of surgical tape appears to give the most accurate results. The average error, taking the thermopile values as standard, was only 0·5–0·8° C, and the thermo-junction temperatures were consistently higher than the thermopile values by about this amount.

The Benedict type of junction also gives results which compare reasonably well with those obtained from radiation measurements. In our observations the average error with this type of junction was 0·7–1·3° C, and the error usually had the effect of giving too high a value for the thermo-j unction temperature. Other observers, however, have measured temperatures with this type of junction which appear to be somewhat low, and it appears possible that the temperature of the observer's hand may have influenced the readings obtained.

Forms of apparatus in which the thermo-junction and its leads are exposed to the effects of the air are liable to give very erroneous estimates of skin temperature. In our own comparisons the skin temperature was underestimated by 3–3·5° C. when the Strauss thermo-junction was used; and examples have been drawn from the literature which suggest that, with other apparatus in which the junction was exposed, the temperature was underestimated. If such forms of apparatus are used for measuring skin temperatures beneath the clothing, they may yield fairly reliable results, but in exposed positions considerable errors may arise. This liability to error is particularly serious where there is a wide range of air temperature and of air movement.

Skin temperatures can be measured with considerable accuracy by means of an electrical resistance thermometer.

In previous studies of the bacterial flora of the upper respiratory tract and its variations during attacks of acute coryza (Hoyle, 1932), it was found that there was in many cases a striking relationship between the appearance of certain organisms with definite pathogenic properties towards the lower animals, in the upper respiratory tract, and the occurrence of acute infections. This relationship was especially marked in the case of the influenza bacillus, and it was suggested that this organism played a prominent part in the aetiology of upper respiratory infections. The significance of the influenza bacillus in acute coryza has been noted by various workers (Noble, Fisher and Brainard, 1928; Burky and Smillie, 1929; Webster and Clow, 1932; Kneeland and Dawes, 1932), but on the other hand various observers have found the incidence of influenza bacilli in the respiratory tract to be the same in normal health and in acute coryza, and Fleming (1929) and Fleming and Maclean (1930) have devised a technique by means of which they claim to have isolated influenza bacilli in 100 per cent, of normal throats. One possible explanation of these conflicting results may be found in the absence of any very exact criterion of what constitutes a typical influenza bacillus.

Muir and Browning (1909) were the first to show that alexin is retained by the filter when fresh serum is passed through a Berkefeld candle. Several filtrations are generally required to remove all traces of the alexic activity, although when a fresh candle is employed for each filtration, the third or fourth filtrate is usually devoid of alexic power (Schmidt, 1914). It is not established whether or not the failure of later filtrates to haemolyse sensitised erythrocytes results from the removal by the filter of some one or of all of the known components of alexin from the serum. We have, therefore, studied the effect of Berkefeld filtration upon the alexic activity of fresh human serum with respect to the thermolabile elements, and the thermostable third and fourth components (Whitehead, Gordon and Wormall, 1925; Gordon, Whitehead and Wormall, 1926).

1. An account is given of sporadic Salmonella infections. In the period 1929–32, thirty outbreaks occurred, involving forty-six cases, whereas, in the year 1933, thirty outbreaks involving forty-four cases were encountered.

2. The investigation has shown that the most frequent cause of such infections are Salmonella types aertrycke, thompson, dublin, enteritidis and european suipestifer.

3. It has been found that the dublin and suipestifer types are definitely more tissue invasive than the other types. Septicaemia and meningitis have been associated with infections due to S. dublin, and septicaemia in cases infected with S. suipestifer.

4. The biochemical serological characters of a new Salmonella type—S. aberdeen—have been described.

5. An infection due to S. potsdam has been encountered. This type has not been previously described as occurring in this country.

1. Eighty-seven specimens of septic pus obtained from patients suffering from recent and old infections were examined fully bacteriologically and cytologically and were sterilised at 60° C. for 2 or 4 hours. The heated samples of pus were used for our experiments.

2. Glycerine extracts were prepared of the pus by two methods referred to in detail in the text.

3. The digestive action of the samples of sterile pus and of their glycerine extracts were tested on inspissated blood serum, gelatin, milk, egg and hydrocele saline. Digestion occurred on all these media in varying degrees with the specimens of pus, but glycerine extracts were more active.

4. It was found that the addition of pus or glycerine extracts of pus with various strains of Bacillus typhosus, B. paratyphosus A, B and C, B. gaertner and B. shiga to milk, produced clotting of the medium at 37° C. in the majority of our experiments.

5. If pus was heated at 60° C. for as long as 6 hours its digestive manifestations were not affected, but 80° C. for 20 min. or 100° C. for 10 min. completely destroyed the digestive activity.

6. At 22° C. the digestive action of pus was weakly manifested, while in the ice chest no reaction whatever took place.

7. Clotting of the medium occurred when pus was added to killed growths of B. typhosus and B. paratyphosus A in milk, but not with B. paratyphosus B or B. gaertner.

8. Human milk did not clot with pus and these organisms.

9. A large number of observations were made with serum exudates from pus, either after filtering through a Seitz E.K. filter or after heating at 60° C. It was found that the filtered or heated fluids might have a similar or a weaker action as compared with pus, or they might be inhibitory.

10. The white cells from several cases of chronic myeloid leukaemia, chronic lymphatic, the myeloblastic termination of chronic myeloid and acute leukaemia, were examined in the same manner as pus. The cells of chronic myeloid leukaemia acted in every way like pus, those of chronic lymphatic and their glycerine extracts were quite inactive; those of acute leukaemia corresponded to chronic lymphatic, while those of a myeloblastic termination gave a weak digestive reaction.

11. Specimens of blood serum from normal human beings and from patients with acute infections, and some specimens of body fluids before and after filtering, were found to inhibit the clotting of milk inoculated with pus and the microbes.

12. Twenty-two specimens of tuberculous pus obtained from abscesses and from pyo-pneumothorax, and seven samples of tuberculous sputum, were examined fully, but with special reference to their cytology.

In eight cases the pus showed no digestive action and no clotting of milk with the microbes, in nine cases a feeble reaction occurred, and in five cases the reactions were like those of pyogenic pus. Thus seventeen of the specimens could bedistinguished readily from pyogenic pus; in the five positive specimens polymorphs were numerous, but one of these cases was complicated by a secondary infection. In every sample of tuberculous sputum polymorphs and tubercle bacilli were numerous and the reactions corresponded to those of pyogenic pus.

A series of nine cases of paratyphoid fever due to infection with anaerogenic strains of B. paratyphosus B are recorded.

Biological examination of these strains shows no difference from the ordinary type of B. paratyphosus B except the inability to produce gas from any “sugar.”

Serologically these strains are identical with the aerogenic type of the organism.

These anaerogenic strains retain their inability to produce gas from “sugar” after subculture on artificial media for a considerable period, and cultivation on a number of different fluid media failed to restore the gas producing power.

In the organic salt fermentation test B. paratyphosus B gives a characteristic reaction which is also given by the anaerogenic strains.

The fermentative varieties of B. paratyphosus B defined by Kristensen and Bojlèn are definite and can be identified as a matter of practical routine, but the test is delicate and requires care in the preparation of the media and especially in the Bitter test.

The existence of these apparently stable fermentative varieties of B. paratyphosus B is likely to be of use in epidemiological investigations.

We wish to thank a number of people whose assistance has made it possible for us to carry out this work. Dr J. A. Charles, Medical Officer of Health, Newcastle-on-Tyne, for permission to publish these cases; Dr E. F. Dawson-Walker, Assistant M.O.H., for histories and specimens from the cases. Dr W. M. Scott for many cultures and other assistance.

Dr M. Kristensen, Copenhagen, Dr M. A. C. Buckell, Bradford, Dr P. L. Sutherland, Wakefield, for paratyphoid cultures and Dr J. W. H. Morrison, M.O.H., Blaydon-on-Tyne, for a number of specimens.

1. In Chatham naval hospital a carrier rate for agglutinable meningococci of 50 per cent, persisted for over a year, during which period there were no cases of cerebro-spinal meningitis among the naval and military population of the port.

2. During the same period the carrier rate at Portsmouth among the contacts to six cases of meningitis was 5 per cent.

3. During the year previous to the present Chatham investigation the carrier rate at Chatham was 13 per cent, among the contacts to five cases of C.S.M.

4. There was no association of the meningococcus carrier rate with density of population or season of the year.

5. The most senior ratings, with the most spacious accommodation, had as high a carrier rate (60 per cent.) as the recruits with the worst sleeping quarters. Ratings of intermediate seniority had an intermediate rate (38 per cent.).

6. Agglutination tests indicated that the antigenic composition of the bacterial herd of N. meningitidis changed rather suddenly, while the total meningococcal carrier rate remained constant.

7. Many of the above recent observations are at variance to what was seen at Chatham, and elsewhere, during the Great War. An attempt is made therefore to reconcile the past with the present experiences.

Our thanks are due to Surgeon-Commander J. A. O'Flynn, who kindly tested our culture medium and supplied the Portsmouth C.S.M. data, and also to Chief Petty Officer R. Kircher, R.N., Sick Berth Staff, who did all the nasopharyngeal swabbing (in order to keep the personal equation constant), as well as much of the technical laboratory work.

1. Provided that due precautions are taken to ascertain the normal level of agglutinins in a given host, and that antigenically smooth strains are used, there is every reason to believe that the occurrence of agglutinins in a titre above the normal range of variation is due to infection—latent, active or past—with the specific organism in question, or in a few instances with an organism, usually of the same genus, sharing a similar antigen.

2. This conclusion may have to be modified for the occurrence of agglutinins to Proteus OX19 in human typhus sera, since the exact relationship of this organism to Rickettsia prowazeki is still doubtful. If, however, future work shows the truth of Felix's (1933) contention that serological types of Proteus X correspond to serological types of typhus virus (Rickettsia), even this apparent exception will fall within the general rule.

3. The examination of sixty-four horse and fifty-three cattle sera, and the performance of numerous cross-agglutination and cross-absorption experiments with rabbit and horse antisera, lend no support to the suggestion made by certain workers of the existence of an antigenic relationship between Brucella strains on the one hand, and strains of Pfeifferella, Pasteurella and Proteus X on the other.

1. Anaerogenic races of B. paratyphosus B are stable in ordinary artificial culture.

2. Aerogenic variants of anaerogenic strains can be produced (though the anaerogenic strain persists) by growth in dextrose or mannitol with a relatively large amount of sodium formate. Such variants can then produce gas from all the other “sugars” fermentable by B. paratyphosus B.

3. Failure to produce gas from sugars by anaerogenic races is due to deficiency in the power of breaking down the formate produced in the first stage of fermentation.

4. A mixture of dextrose and formate in peptone water reveals the latent capacity for gas production in anaerogenic races of B. paratyphosus B. This confirms the observations of Grey (1914, 1920, 1924).

Since this work was done, we have seen the paper of Pot and Tasman (1932) in which they succeed in isolating aerogenic variants from anaerogenic cultures of B. paratyphosus B by continued cultivation in 2 per cent. calcium formate, in a somewhat similar manner as we have done with the mixture of dextrose and sodium formate.

Autopsies on all the guinea-pigs kept for as long as 20 months failed to reveal any trace of tuberculosis, although the grass grazed by the animals was proved by subcutaneous inoculation to bear virulent organisms, which might have been expected to reach the alimentary tract in considerable numbers, at least in the earlier experiments.

By repeated infection of a grazing area it has been possible to induce tuberculosis in guinea-pigs grazed in the open and in those fed indoors on cut grass. In the 1932 experiment a total of thirty guinea-pigs were allowed access to plots of grass which had been infected, one, four and eight days previously and all failed to become infected. The combined experiments (1932 and 1933) suggest that under some weather conditions, animals, although admitted to pasture within a very short time of its infection may escape infection with tuberculosis, but that repeated infection of pasture results in a high incidence of disease in animals grazing on it.

Culbertson (1932) has recently described a method for estimating the antibody content of antisera against crystalline egg-albumin. He determined by tests on supernatant fluids the point at which antigen and antibody neutralised each other. At this neutralisation point he found that antigen and antibody combined in a ratio of 1 to 13, and that the amount of precipitate nitrogen was maximal. The ratio 1 to 13, he stated, held for all the precipitates in the post-zone where antibody was in excess.

The haemolytic streptococci associated with scarlet fever, tonsillitis, septic conditions, etc., belong to one group or species designated Streptococcus pyogenes.

Cultural and serological aids to the identification of Str. pyogenes are described.

The epidemiologically significant types of Str. pyogenes appear to be about 20 in number, though probably more than thirty type sera will be necessary for the complete analysis of the group. So far twenty-seven individual serological types have been defined.

The irregularities which have been observed in the agglutination reactions of streptococcal cultures are ascribed chiefly to variations in type-specificity.

Different colonies may give “type-specific” or “group” agglutination, and these distinctions are revealed macroscopically in the case of certain types by growth on homologous agglutinating serum agar.

Some examples are given of observations on the epidemiology of streptococcal infections in schools and other institutions.

The principles involved in the serological classification of a bacterial group are discussed.

The importance of pulmonary tuberculosis as a cause of death in Wales has been the subject of much consideration. Many of the investigations, however, have been of a definitely local character, though this fact does not detract from their merit as is evinced by the work done by Wade, Bowen and the King Edward VII Welsh National Memorial Association. For a really comprehensive study we are indebted to the late Dr Brownlee. His contribution to the epidemiology of phthisis was unique in its suggestiveness. He postulated the theory that phthisis as a disease was, not an entity, but that it consisted of three types—“Young Adult,” “Middle Age,” and “ Old Age,” corresponding with separate strains of the infecting bacillus. According to his conception the great bulk of the phthisis deaths in Wales was due to early or late types of the disease. He reached this conclusion from a study of the pulmonary death-ratesof males between 1851 and 1870 in the registration districts of Wales.

The Isle of Ely is the northern half of the geographical County of Cambridge-shire and forms in itself a separate administrative county. It is the heart of the fen country, and its population is one which long retained a degree of isolation and independence tending to preserve in its members certain mental, and perhaps physical characters peculiar to themselves. It is interesting to recall that from this neighbourhood came Boadicea, Queen of the Iceni, and Hereward the Wake, who proved such stubborn enemies to the conquerors of their country.

In 1928–9 was published the result of an investigation into the incidence of tumours of the respiratory tract based on an analysis of the post-mortem records of the General Infirmary at Leeds between the years 1891 and 1927 inclusive. It was shown that there had been no increase in intrathoracic cancer at post-mortem during this period, that the incidence was not as high in Leeds as it was in Manchester and that the disease affected men more commonly than women in the proportion of 3½: 1. It was also shown that there was a general tendency towards an increase in lung cancer in post-mortem statistics of other hospitals, especially in Germany.

The serious problem presented by the high prevalence of severe pulmonary tuberculosis in early adult life in the Highlands and Islands of Scotland is exemplified by the figures in the following table, for which we are indebted to the Department of Health and the Registrar-General for Scotland.

By feeding monkeys on healthy rye, wheat and other cereal and leguminous seeds convulsive and paralytic symptoms similar to those of convulsive ergotism in man can be produced.

Large amounts of cold-water extracts of the grains given to monkeys per os may cause the symptoms acutely, just as large meals of rye bread have occasionally been reported to do in man.

Salts of phytic acid and decomposition products from it, isolated from all these grains and given to monkeys by the stomach or hypodermically, occassion symptoms exactly similar to those caused by feeding the grains.

The occurrence of poisoning when these grains are consumed as food is partly a question of the quantity consumed and partly a question of the ability of the consumer to break down in the bowel the poisonous phytates and so render them innocuous. If they are not fully broken down they are absorbed and act as poisons to the nervouse system.

The pathological lesions in the nervous system of monkeys are the same as those which are described as occurring in convulsive ergotism in man.

Convulsive ergotism is not a “deficiency” disease, nor is it ergot disease, but is caused by poisons normally present in rye and other grains.

1. This paper deals with the epidemiology of an outbreak of influenza which started in Leeds late in November 1932 and continued until the end of February 1933.

2. Returns of the actual cases which occurred each week in certain official departments and large firms made possible the investigation of a selected population. This consisted of 17,195 persons, and among its members there were 2772 cases of influenza with five deaths. The case-rate was therefore 16·1±0·2 per 100 of the selected population. It is shown that in this popu lation factory workers were less liable to contract influenza than were assistants in large shops, outdoor workers, clerks and office workers, and school teachers; and it is suggested that propaganda relating to respiratory infections may have had some effect in lessening the incidence of the disease among the workers in these factories. The mean duration of the outbreak in the selected establishments was 11·55±0·29 weeks.

3. The total cases which occurred each week in the selected population were plotted as a graph; the resulting curve was of the typical epidemic type. A Pearson's Type IV curve was fitted to these data. The available evidence points to the fact that the outbreak consisted of a single wave.

4. From the data for the city as a whole it is shown that the Leeds out break took approximately twice as long to reach its peak as did the Glasgow epidemic, though both outbreaks started practically simultaneously. Never theless, in actual epidemicity, the Leeds outbreak did not fall far short of that which occurred in Glasgow.

5. An investigation of the age distribution of deaths from influenza in Leeds since 1900 shows that, though the 1918 epidemic produced a sudden change in the age distribution, in that there was an unprecedented toll of young adult life, the return to the status quo was not sudden, as is often assumed, but was effected gradually over a period of about ten years.

6. It is shown from an investigation of influenza deaths in different parts of the city that there was no direct association between these deaths and the density of population in the wards in which they occurred. Neither was there any association between the number of influenza deaths and the average number of persons per room in each ward. From a study of the distribution of the deaths in time and space it appeared that, with certain exceptions discussed in the paper, the outbreak started in the centre of the city and extended outwards from the centre with decreasing potential.

7. An investigation of the meteorological conditions suggested that there was a definite relationship between a fall of temperature, especially if accompanied by fog, and an increase in the incidence of influenza and of pneumonia.

8. The clinical features of the outbreak are discussed fully in the text. The disease was essentially of the respiratory type. Two forms of this were in evidence, (a) a form resembling the classical picture of influenza, and (b) a form in which the predominant feature was tracheitis. The gastric type of influenza was not common. Complications were not very frequent, but the commonest were influenzal pneumonia and acute otitis media. The duration of incapacity for work varied according to the social status of the individuals. Brain workers and persons in the middle classes were generally off duty for about nine days. On the other hand, the duration of incapacity in artisans and factory workers was usually about a fortnight or three weeks.

9. From the evidence which is adduced in this paper as a whole, including a consideration of the tuberculosis statistics of the city, it would seem that, though the outbreak of influenza in Leeds during the winter of 1932–3 was of a mild and non-killing type, infection was widespread, and the total incapacity must have resulted in a considerable economic loss to the community.

I am indebted to Prof. J. Johnstone Jervis for suggesting this investigation. My thanks are also due to the heads of the various departments and commercial firms, and to the Staff of the Public Health Department, without whose co-operation the data for this study could not have been obtained.