1. The physical, climatic and economic conditions in Nairobi township are briefly discussed. The factor which appears to be most closely related with the incidence of plague is rainfall. A growing season producing, mainly by virtue of ample rainfall, large crops of cereals, which for various reasons may be kept in storage for long periods and become easily available to the rat population, is invariably followed by plague.

2. The numbers of Indians and Africans are given together with the incidence of the disease in these two types and the localities where the disease originates and persists.

3. The distribution of rat and flea species is outlined and shows an irregular distribution of the species of Xenopsylla. The rat population during outbreaks shows a higher density than during quiescent periods, but not such as to indicate that numerical strength alone is the main factor causing outbreaks of plague in the Rattus population. Higher flea populations follow high densities in the rat population.

4. Attention is drawn to the area where plague originates and persists and to the mixed flea population of the locality. Pumwani and Pangani with almost pure cheopis colonies do not figure in early outbreaks and in them the disease ceases much earlier, but they suffer heavily during epidemics.

Outbreaks of plague in Nairobi are associated with grain storage areas and negligence and improvidence are important contributory factors.

5. The results of the survey have enabled some advance to be made in understanding the distribution of the two main species of Xenopsylla. All Rattus living in earth burrows have been found infested with Xenopsylla cheopis, and Rattus living in roofs or walls with Xenopsylla brasiliensis. These facts explain why such places as Pumwani with its African population has a cheopis colony and the K.A.R. lines have brasiliensis. In the Indian Bazaar with varied facilities for nesting, there is a mixed population of fleas.

The application of these facts to rural areas explains why brasiliensis is almost the sole ectoparasite of Rattus living in thatched roof habitations, and why field rodents such as Mastomys coucha and Arvicanthis abyssinicus living underground are hosts to Xenopsylla cheopis.

1. In the introduction, the characteristics of the people are given from the observations of administrative and medical officers well acquainted with the area, fertility of the land and ample rainfall being the main theme of each observation.

2. The topography, climate and local conditions are given, including notes on crop production, habitations, clothing, domestic animals, origin of the people and food.

3. The population figures and rat and human plague figures are discussed. The rat and flea population figures over a twelve month trapping period are given and notes made on interesting biological problems arising out of these figures.

4. The male and female ratios in a Rattus population in plague endemic areas offer an interesting problem for further investigation. It would be instructive to find out what are the factors responsible for high female ratios and the significance of such female dominance being peculiar to endemic areas.

5. The field rodent and flea survey figures are discussed. Although an interchange of fleas takes place between them and house rats, plague remains absent from such communities.

6. There is no evidence to show that there is carriage of infection from endemic centres, either by rats, fleas, merchandise or man.

7. A comparison is drawn between a Kenya endemic centre and an Indian one, similarities and differences being noted. The similarities are that both areas are extremely fertile, large crops of cereals being grown, and the diets of both man and rats are confined to this inadequate source of nourishment. The great differences between Indian and Kenyan endemic centres are the flea species present and that India suffers a much heavier mortality rate in man than does Kenya.

8. It is suggested that the low incidence of plague in man in Kenya is partly explained by the habits of Rattus, which in the typical dwellings of native areas nests in the thatch. Its ectoparasites are then unlikely to come into contact with man. In buildings where rats nest underground contact between its flea (then Xenopsylla cheopis) and man is likely to be very close and constant, and areas containing such buildings are usually distinguished by the epidemic nature of the disease.

1. Nairobi township and other areas about the Mount Kenya endemic centre have frequently been penetrated by persons suffering from plague, but, at least during the last few years, plague has failed to become established in this way.

2. A survey of rats and fleas, conducted at Mombasa upon 66 truckloads of maize in transit, resulted in no rats being seen or captured, whilst only seven fleas (2 ♂ and 2 ♀ X. cheopis, 1 ♂ and 1 ♀ D. lypusus and 1 ♀ Ct. cabirus) were collected from clean white rats after they had been allowed to run about in the maize.

3. Although maize continues to be repeatedly unloaded and handled at Mombasa, this town has remained free from plague.

4. We have failed to obtain any evidence that the transport of maize from endemic centres by railway is a factor in the dissemination, of plague.

A simple process is described for the desiccation of antisera after freezing by rapid evaporation. Freezing has been brought about by the rapid evaporation which takes place in a high vacuum in the presence of P2O5. The violent frothing which takes place when serum is exposed to a high vacuum has been controlled by a preliminary evacuation of the serum in the absence of the drying agent.

No special apparatus except an efficient high vacuum pump is necessary for this process.

This process has been applied for the preservation of (a) anti-horse serum, anti-ox serum, anti-human serum, anti-human globulin and anti-crystalline human albumin sera; (b) anti-sheep cell haemolysin; (c) guinea-pig complement; (d) various Salmonella antisera.

Preliminary experiments are described of the application of this method to the drying of bacteria.

We desire to thank Dr P. Hartley for his valuable advice and for showing us his paper prior to publication.

We have been able to confirm all the experimental results of Felix and his collaborators in regard to the Vi antigen and its significance in mouse infections. In the interpretation of some of the findings we are, however, not quite in agreement with Felix.

When large doses of living V forms of typhoid bacilli are injected into the mouse they cause acute intoxication followed by death within 2 days—if the animals die. Surviving animals are clinically healthy after this time, often earlier, but they are nevertheless found to be massively infected in the organs and blood for several days.

When smaller, non-intoxicating, doses are used it is not possible to demonstrate any difference between the V form and the W form of the same strain in regard to the severity of the infection they cause; so that we may conclude that the Vi antigen plays no part in the virulence of the typhoid bacillus in the mouse. (Remember in this connexion that the dysentery bacillus is also very toxic for mice injected in large doses, while its virulence, i.e. the ability to grow in the mouse is = 0; but when rapidly fatal doses are used a massive growth of dysentery bacilli can also be demonstrated here, the cause being that the natural immunity has become partly neutralized by the intoxication.)

Like Felix we find that only the V form can be used both for active vaccination and the production of sera which protect against lethal doses of V forms in mouse experiments. We consider that this is due to the antitoxic effect. Our experiments show that the anti-bacterial effect in vaccination experiments differs little whether the V or the W form is used as vaccine, so long as the vaccinal effect is tested with non-toxic doses of bacilli. As shown in earlier experiments, the bactericidal effect in actively vaccinated animals is often a relatively slow process.

Only the Vi antibody containing sera can protect prophylactically with certainty against the toxic effect of living typhoid bacilli. On the other hand; the bactericidal effect prophylactically of 0 and 0 + Vi sera is very similar when tested against non-toxic doses of “virulent” typhoid bacilli.

The therapeutic value of potent Vi sera under the experimental conditions given was nil.

Salmonella infection is fairly frequent in some animals, but is not very common in sheep. Moreover, food poisoning in man associated with mutton is comparatively rare. It is true that S. typhi-murium is sometimes referred to as the “mutton type” of Salmonella (Schütze, 1920; Lovell, 1932 b).

The filtration end-point of 0·2μ leads us to assign according to Elford (1933) a value 100—150 mμ. for the particle diameter of the virus of rabies. This figure we have previously reported in Ann. Report Med. Res. Council (1934) and also in a discussion held by the Royal Society of Medicine (Galloway, 1936). Recently the Japanese workers, Yaoi et al. (1936), using similar graded collodion membranes (gradocol) and the same general technique as employed in the present experiments, have studied independently another strain of “fixed” rabies virus “Fukuoka”. Their stock filtrates appear to have been more potent than ours, their figures for the limiting infective dilution being 10–3 to 10–4. The incubation period in rabbits inoculated with the “Fukuoka” strain is also shorter than in the case of rabbits inoculated with Pasteur Institute (Paris) strain. Even so, they find exactly the same value for the filtration end-point 0·2 μ as found by us for the Pasteur strain, thus indicating the particle size of these two “fixed” strains of rabies virus to be the same.

The virus of Aujeszky's disease (pseudo-rabies) has been found to have a particle size of 100—150 mμ. Two strains were employed in these experiments, one that of Aujeszky, and the other Shope's “mad-itch” strain. The results with both strains were identical.

In an article on Pneumonoconiosis published in 1935 I attempted to simplify the problem of lung diseases caused by exposure to dust in industry by means of a classification which might allow of the placing of individual cases into appropriate categories. The importance of some such classification has been emphasized during the past year or so by the gradual realization of the fact that, especially in the South Wales coalfield, cases of more or less disabling lung conditions are frequently met with which cannot conscientiously be fitted into the category of “silicosis”, as defined at the International Conference at Johannesburg in 1930. These cases cannot be “certified” for compensation under existing Statutes, and yet present a picture of respiratory distress and dyspnoea on exertion which marks them out as quite unfit for work.

Our knowledge of the antigens of the typhoid bacillus has recently been enlarged by the investigations of Felix & Pitt (1934).

It has been recognized that the Vi antigen, newly established by these workers, is not only a very important factor in immunity, but also apparently a more labile substance than the O antigen hitherto regarded as the portion of the typhoid soma vitally concerned with the production of prophylactic immunity.

Two batches of pooled antisera obtained from rabbits injected with crystalline egg albumin have been fractionated with ammonium sulphate in order to investigate the distribution of the antibody to this antigen in the serum proteins.

The albumin fractions contained no antibody.

All the globulin fractions examined contained antibody, and flocculated readily when mixed with appropriate dilutions of the antigen.

The distribution of antibody throughout the serum globulin was fairly uniform, the ratio of the antibody content to the protein concentration being the smallest in a fraction with a solubility in ammonium sulphate slightly lower than that of pseudoglobulin.

1. Our experiments show that an intravenous injection of antigen immediately after an intravenous injection of antiserum in the guinea-pig were followed by:

(a) Acute shock and death within 5 min. The signs and post-mortem appearances were indistinguishable from acute anaphylactic shock as typically seen in the guinea-pig. Some control animals injected with antiserum only died in a way similar to those which received both antiserum and antigen. These are discussed in detail in the body of the paper.

(b) Delayed shock and death some hours later. The post-mortem appearances were those of gastro-intestinal congestion and haemorrhage resembling the changes seen in dogs dying of anaphylactic shock. Such changes were never seen in the control animals.

(c) Recovery. Practically all the animals which recovered had symptoms of respiratory embarrassment immediately following the injections of anti-serum and antigen and many had later symptoms of abdominal shock. The animals which were given an injection of antiserum only rarely had any symptoms and never abdominal symptoms.

2. It is necessary to test the antisera used by control inoculation since some antisera are toxic.

Conditions are described under which late death from active anaphylaxis occurred in the guinea-pig. The post-mortem appearances were those of gastrointestinal congestion and haemorrhage resembling the changes seen in dogs dying of anaphylactic shock.

Late deaths can be produced by the intravenous injection of antigen but more easily by intraperitoneal injections.

In a previous paper (Maddock, 1934), it has been shown that healthy calves may be infected with bovine tubercle bacilli when grazed upon pasture infected at intervals with emulsions of the virulent organisms. These infections were designedly heavy, and it was felt that this orientating experiment should be followed by one in which pasture infection was secured in a way more closely approximating to the natural.

Physiologists have recognized for a long time that stuffiness in a close atmosphere is not due to diminution of oxygen or excess of carbon dioxide in the air breathed, or to any other impurity in the air. Such feelings are now shown to be due to the effect of longer infra-red rays on the skin, and are accompanied by a reflex narrowing of the air-ways of the nose and lungs. These effects are set aside by cool air acting on the skin and in many people also by shorter infra-red rays from bright sources of heat. This is why some people are much more comfortable with a bright source of heat than with a dark or dull red one. When people go out of doors to get a “breather”, or stand before an open window during physical exercises, they are instinctively using the stimulus of cold to widen their air-tubes. In open air treatment, and when sleeping with an open window, the same effect is obtained.

All rooms should be adequately ventilated with cool air to prevent stuffy feelings, and keep the air-ways open. Such ventilation will, at the same time, sweep away infection spread by “carriers”.

Partial obstruction of the air-tubes hinders the passage of blood through the lungs; it is this which causes failure of breathing when the obstruction is considerable.

Bleeding is of value because it relieves to a certain extent the impeded circulation.

Breathing through a narrow tube can be used as a test of fitness. An airway gauge has been designed for making such a test.