1. In Hong Kong, where scarlet fever is extremely rare, the Dick reactions of a random sample of 921 Chinese were ascertained. Positive reactions were obtained in 19·1 per cent. The distribution of positive reactors in different age groups is recorded; it bears a general similarity to those recorded by various observers in Europe and the U.S.A.

2. A haemolytic streptococcal carrier rate of 6·6 per cent, was determined by nasopharyngeal swabbing of 1600 Hong Kong Chinese. The frequency of haemolytic streptococci was higher during the cooler months of the year and among the younger age groups.

3. Erythrogenic toxin production and fibrinolytic activity were demonstrated in a number of the strains of streptococci isolated from apparently normal throats.

4. The significance of these findings is discussed.

1. An antiserum containing antibodies for two unrelated antigens, crystalline egg albumin and crystalline horse serum albumin, has been prepared.

2. In titrations of the antiserum against mixtures of these antigens it has been possible to produce a single zone or two zones of optimal particulation by varying the relative concentrations of the components in the antigenic mixtures. It is evident, therefore, that whilst (a) a single zone does not necessarily indicate a single antigen-antibody system, (b) multiple zones suggest the presence of more than one antigen-antibody system.

1. By titration against horse serum and its protein fractions anti-horse sera can be divided into three groups, according to the antibodies responsible for the main or earliest particulating zone, namely the sera containing antibodies against (a) globulin only, (b) albumin only, and (c) both globulin and albumin.

2. For the quantitative estimation of a protein fraction in a horse serum it is necessary to use an antiserum containing a main zone antibody acting with that protein only.

This is illustrated in experiments, designed to show the increased globulin content of diphtheria antitoxin, in which antisera containing antibodies against both globulin and albumin failed to register an increase in the globulin content.

A strain of Trypanosoma brucei when recently introduced into mice and imperfectly accommodated to this species of host produced infections relatively resistant to various trypanocidal drugs. On the other hand, when the strain had become highly accommodated and its pathogenicity increased to a maximum as the result of repeated passages, infected animals were readily cured. The exact mechanism on which this difference in resistance may depend has not been investigated. Attention is directed, however, to the important fact that the drugs acted poorly when the strain was in an attenuated state, as shown by the prolonged course of infection and the occurrence of marked fluctuations in the number of parasites present in the blood. Thus the chemotherapeutic response was weak at the time when the host itself was able to exercise an effective resistance. Later, when the host's resistance had become negligible, the curative action of the drugs was pronounced.

For the investigation of the Hill phenomenon, viz. the antagonism of short-and long-wave infra-red radiation in their influence on nasal obstruction, the commonly used rhinomanometric technique is not well suited. In this paper a new procedure is described, in which an air current is blown through the nose. In this way the nasal passability may be estimated, independent of respiratory movements. It proved to be desirable to make a selection of suitable experimental subjects with the aid of adrenaline and histamine tests; only those persons in which the opening by adrenaline and the narrowing by histamine sprays were clearly demonstrable were chosen.

Long-wave infra-red rays constantly caused nasal obstruction.

The shorter waves were much weaker in their shutting effect, and also decreased the narrowing of the nose when they were given in combination with the long-wave rays of an electric fire.

A nasal passage closed by long-wave infra-red radiation may be partly opened by the short-wave rays; the possibility of surface cooling being the opening factor was excluded.

Artificial cooling of the skin has a very marked opening influence.

It is suggested that nose-opening and nose-shutting are correlated with cooling and heating of the Malpighian layer. Long infra-red rays will be completely absorbed in the layer and cause capillary stasis with local overheating. The shorter waves for the greater part penetrate more deeply; the skin reacts by an active hyperaemia, which, by relatively cooling the overheated Malpighian layer, has the same effect as a cooling from without.

Investigations on perspiration, particularly in regard to feet, are not easy to carry out; observations have to be made under a variety of conditions and complicating factors. A man's foot perspiration may be alkaline whilst another's is acid, and the same person may show different reactions on the same day.

Some months ago Prof. M. J. Stewart (1933) gave an address on silicosis in which he covered the whole of the pathology of the various types of the disease. It may seem strange, therefore, that the same subject should be chosen for discussion so shortly afterwards. My apologia is that silicosis presents manysided problems and that since Prof. Stewart's paper appeared the claim has been ventured that the cause of silicosis has been demonstrated. Discussions have been taking place during the past twelve months at the Institution of Mining and Metallurgy—discussions stimulating in themselves and showing the necessity of further research along old and new avenues.

It is now almost universally recognised that a test for disinfectants which is to yield results of direct practical significance must include appropriate “organic matter” in the test mixture. Some types of disinfectant, notably those which act by liberating chlorine or oxygen, are greatly reduced in activity by the addition of only small quantities of any extraneous material with which these elements can combine. Other disinfectants are more resistant to this influence; they are capable of efficient action in the presence of excreta and dirt of other kinds if used in adequate concentration. What this concentration should be under such unfavourable conditions it is necessary to determine, since the degree of interference is not the same for all disinfectants even in the same class, and no constant relationship exists between efficiency under these conditions and that displayed in a medium of distilled water as determined by the Rideal-Walker test. Numerous efforts have been made in the past to introduce tests employing adequate amounts of complex organic matter; few of these have achieved permanence and none popularity. The following short survey of past work is concerned only with what may be termed “excremental” disinfectants; the testing of surgical disinfectants, which may reasonably be held to demand different methods, is outside the scope of this paper.

Fifty-one strains recently isolated from cases of bovine mastitis have been studied in their morphological, cultural and biochemical reactions.

Using their colony appearance in blood agar as the first differential criterion we have shown that the strains differ widely from one another and from the α β and γ streptococci usually encountered in human disease.

The heterogeneity of mastitis streptococci was strikingly demonstrated. From the results of a relatively small number of cultural and biochemical tests it could be shown that no two of the fifty-one strains were identical, in every respect.

The findings recorded suggest that great caution must be exercised in ascribing a bovine or human origin to individual strains of streptococci on the results of cultural or biochemical reactions.

During the experimental comparison of different types of hot-water heating systems in a school at Easter 1934 about a thousand kata readings were taken. These were taken mostly in pairs, one immediately after the other, and a third reading was taken only when the first two differed by more than about 10 per cent.

1. Previous work on the antigenic structure of the cholera group of vibrios is experimentally reviewed, and the data amplified and systematised.

2. The cholera group is defined as consisting of vibrios with similar biochemical characters and having a common heat-labile antigenic component.

3. The heat-stable antigens are divisible into:

(a) A considerable number of specific antigens, best demonstrated by O sera and H-O suspensions, which serve as a basis of classification into O subgroups.

(b) A non-specific component, demonstrable with O sera and O suspensions.

4. The first subgroup contains all the standard cholera vibrios from central laboratories, and the majority of other epidemic strains. We consider that it represents the only class of vibrios known for certain to cause epidemic cholera.

5. The races of this subgroup I are further divisible into two (or perhaps three) “types”, as established by Japanese workers, according to differences in their subsidiary O antigens.

6. The haemolytic “El Tor” vibrios are serologically diverse. The term “El Tor” should, as Shousha suggests, be reserved for those that have the same specific O component as the standard cholera vibrios.

7. For the identification of the undoubted cholera vibrios a standard subgroup I O serum is recommended in conjunction with the haemolytic test. The serum should contain both the main and the subsidiary antigens of the subgroup.

8. As a working rule it is suggested that bacteriological proof of “cholera” or a cholera carrier should rest on the isolation of a non-haemolytic vibrio with the specific O antigen of subgroup I.

The studies and observations on which this paper is based were conducted while one of us (K. V. V.) was holding a Fellowship of the International Health Division of the Rockefeller Foundation.

Our thanks are due to Mr P. Bruce White for constant help and advice, and to the numerous bacteriologists in various lands who have kindly supplied us with cultures and sera.

1. In view of the high carrier rate on discharge from hospital, the bacteriological examination of swabs from scarlet fever patients is of no value in the detection of probable infecting cases.

2. There is, however, a definite relationship between the degree of infection, as measured by the profusion of haemolytic streptococci in cultures on discharge, and the likelihood of the patient infecting others.

3. The causal connection between “infecting” case and return case is strongly supported by finding the same serological type of Streptococcus pyogenes in both cases.

4. A patient who has had complications in hospital is slightly more liable to become an infecting case than one whose course has been uncomplicated.

5. The commonest age group of infecting cases is 5–10 years.

6. The majority (70 per cent.) of the return cases occurred within 14 days of the arrival home of the infecting case.

7. Overcrowding in the home, and the number of the susceptibles exposed do not appear to be important factors in the production of return cases.

8. A history of tonsillectomy does not appear to bear any important relationship to the occurrence of infecting cases, but the inconclusive nature of our findings indicates the desirability of further investigation of this question and that of the relation of the condition of the tonsils to their infectivity.

9. Discharge from hospital as early as is consistent with a satisfactory clinical condition is shown to be advantageous from the point of view both of the patient and of the hospital administration.

10. The administration of scarlatinal antitoxin is likely to render a patient less liable to convey infection on discharge.

1. The tuberculin reactions of 259 Hong Kong Chinese, mostly adults, were determined by the Mantoux method. Each individual was tested with various dilutions of both old tuberculin and tuberculo-protein.

2. Of 215 subjects above the age of twenty, 93·1 per cent, yielded positive reactions to old tuberculin in dilutions of 1/1000 or higher.

3. Tuberculo-protein compared with old tuberculin of similar dilutions did not yield such a high percentage of positive reactions, seldom produced such vigorous reactions and never caused vesiculation. (This may have been due to deterioration.)