We confirm the finding of Behnke, et al. (1935) that air at 8·6 atm. pressure has a somewhat intoxicating effect on human beings, and that this effect is due to nitrogen. The nitrogen effect reaches its maximum after about 3 min. There was no reduction of manual dexterity in the test used by us, but a considerable effect on performance of arithmetic, and on most practical activities. At 10 atm. these effects were somewhat enhanced, and manual dexterity was lowered in some cases. When helium or hydrogen was substituted for nitrogen there was no intoxication.

3–4% of carbon dioxide at atmospheric pressure caused no deterioration in manual or arithmetical skill, and in the two subjects tested, 6% of carbon dioxide caused no deterioration.

When air containing about 0·4% of carbon dioxide, and therefore with a partial pressure of about 4%, was breathed at 10 atm., there was a marked deterioration in manual dexterity, and a good deal of confusion. When breathing carbon dioxide at partial pressures of 6·6–9·7% at 10. atm., eight subjects lost consciousness in 1–5 min., but some could tolerate partial pressures of over 8% for 5 min. or more. With half an hour's exposure to a partial pressure of 6–7% of carbon dioxide, one subject lost consciousness after 7 min. at 10 atm. pressure, and another nearly did so.

We consider that the percentage of carbon dioxide in air at 10 atm. pressure should be kept below 0·3%. Exposure to high partial pressures of carbon dioxide at 10 atm. does not increase the liability to ‘bends’ or other symptoms due to rapid decompression.

Immersion in water below 40° F. did not enhance the effects of high-pressure air, or of carbon dioxide at atmospheric pressure, but somewhat enhanced those of high pressure and carbon dioxide together.

In certain breathing apparatus the resistance became so great at 10 atm. as to be intolerable.

Few subjects experienced serious trouble during compression, or during or after decompression. But one developed a unilateral pneumothorax.

The susceptibility of the field vole to experimental infection with bovine, human and avian bacilli respectively was the subject of a previous paper (Griffith, 1939).

Subcutaneous inoculations: Two strains of human origin were tested, one was of eugonic human type from the sputum of an English case of pulmonary tuberculosis; the other was of dysgonic human type obtained by Dr J. A. Young from a native in Nigeria. Both strains produced typical general tuberculosis in the hamster, the lesions on the whole showing less caseation than those set up by bovine bacilli (vide 1939).

The vole strain of bacillus gave rise to generalized disease resembling tuberculosis. The duration of life of the animal was, however, prolonged and though the bacilli were abundant in the lesions there was no necrosis or caseation. Similar results in two hamsters were reported in 1939.

The avian bacillus was the least pathogenic of the four types. Of the three hamsters inoculated only one (H. 17, dose 5·0 mg.) showed macroscopic foci in the liver and spleen; as these were not seen microscopically in sections it is doubtful if they were due to the action of tubercle bacilli. It is unfortunate that the specimens of Hamster 17, which died in my absence from the laboratory, were badly fixed; a smear preparation of the spleen was not made. In the others tubercle bacilli were found in the organs, but were not numerous. The peculiar result after 5·0 mg. of avian bacilli—enormous multiplication of bacilli in spleen, liver and a local gland without macroscopic tuberculous lesions—described in the first paper was not reproduced.

Feeding: The four different types of bacilli were given by the mouth to six hamsters, the dose in each instance being 10 mg. of culture. Two of the hamsters, one fed with a potato culture of the vole strain, the other with a human strain, escaped infection. The remaining four hamsters, each fed with one of the four types of bacilli, developed disease, the severity of which in each instance was in harmony with that following subcutaneous inoculation of the type.

Intravenously injected into a goat in a single dose of 50·0 mg. b.c.g. caused the death of the animal from pneumonia in 9 days and was demonstrated in the blood until the day of death, though in progressively diminishing numbers.

Intravenously injected into a goat on successive occasions, the dose on each being 10·0 mg., b.c.g. was filtered from the blood by the tissues within 3–5 days. There was little or no evidence that b.c.g. disappeared from the blood more quickly after previous injection of that micro-organism. b.c.g. was not excreted into the milk cisterns of this goat or, if excreted, the bacilli did not long survive.

Intramammarily injected b.c.g. produced inflammatory reactions in the mamma with formation of caseo-pus and ultimate destruction of the gland as a secretory organ. The micro-organisms, however, did not remain alive for any lengthy period in the mamma and in the two instances reported were not viable after about 15 weeks.

Pus from a lumbar abscess in a boy, aged 10 years, produced severe general tuberculosis in a guinea-pig. A culture from this guinea-pig resembled a bovine strain in primary culture on egg and a eugonic human strain in subculture on glycerine egg and glycerinated potato. The virulence of the strain was high for the guinea-pig and irregular for the rabbit. For some rabbits the pathogenic effects of the strain were much more severe than those of typical human bacilli and approached, but did not equal, those of bovine bacilli. The results in other rabbits were similar to those of eugonic human bacilli. Cultures recovered from the rabbits resembled in characters the original strain, which was therefore not a mixture of the human and bovine types of tubercle bacilli. The strain cannot be described as ‘human’ or ‘bovine’ according to the usually accepted distinction between these types in the rabbit, but according to the results in the vole could be classified as belonging to the human type.

The joints of rabbits are predilection sites of disease in chronic avian infections and are nearly always found affected in rabbits which have lived longer than 3–4 months after an intravenous or subcutaneous inoculation of fully virulent avian strains. Highly attenuated avian strains, namely, those which have ceased to be infective for fowls, also give rise to joint disease in rabbits when injected intravenously; for example, twelve rabbits which lived over 5 months after intravenous inoculation of 2·5–10·0 mg. of attenuated avian strains all showed joint tuberculosis, and in eleven of these the joints were the only affected parts (Griffith, 1925).

It has been found that when a toad is injected subcutaneously (e.g. in the dorsal sac) with tubercle bacilli the micro-organisms disseminate but are stored in the liver mainly, where they multiply slowly without producing visible lesions and can be recovered by culture or injection of guinea-pigs after several years. Bovine tubercle bacilli were recovered in culture from the liver 1, 2 and over 3 years after subcutaneous injection and tubercle bacilli of avian type were still viable after 900 days in the liver.

A series of 35 unselected cattle tested with synthetic medium tuberculin and synthetic medium johnin was subjected to post-mortem examination. In five instances (14·3 %) avian strains of full virulence for the usual experimental animals were recovered by the inoculation of guinea-pigs with glandular suspensions or by cultural methods used for the isolation of Johne's bacillus: in two of these cases an avian type was isolated from the glands of the alimentary tract and a bovine type from the thoracic glands. In addition there were five cases (14·3%) in which evidence of avian infection was not complete. From one a strain was obtained which was culturally identical with an avian type, but it was almost completely avirulent for laboratory animals. In two, acid-fast bacilli were seen in smears from the thoracic and mesenteric glands respectively, while in two, completely calcified glands were observed which were typically tuberculous, but were devoid of acid-fast bacilli (microscopical examination of smears and sections). Cultures from the organs of guinea-pigs inoculated with glandular suspensions from the four last-mentioned cases failed to grow tubercle bacilli. In these circumstances the evidence that the infection was of the avian type is largely presumptive, but the nature and extent of the lesions suggests that such may have been the case.

With, regard to the tuberculin and johnin tests, clear cut reactions to johnin were obtained in seven instances, a doubtful reaction in one and negative reactions in two. One-half of the animals were also positive to tuberculin.

In the last century a wasting disease of cattle in which diarrhoea was the prominent symptom was described by some veterinarians, but was at first often confused with tuberculosis although the distinctive lesions found in the latter disease were not present at post-mortem examination. However, it is only during the last forty years that Johne's disease has been clearly differentiated from bovine tuberculosis.

Over a period of 6 months, wounds and throats in a plastic surgery ward were swabbed. Haemolytic streptococci obtained from them were typed by the late Dr F. Griffith and his staff at the M.E.C. Streptococcal Research Laboratory. 65% or more of the wounds in the ward were found to be infected with Streptococcus pyogenes, but relatively few suffered severe damage as a result of their infection.

About a quarter of the cases admitted to the ward acquired streptococcal wound infections while in the ward. Two wound and throat epidemics, one due to type 13 and one to type 4/24 streptococci are described.

1. The agglutination titres for sheep and for horse erythrocytes of 100 normal sera and twenty-seven glandular fever sera have been determined.

2. The results of the sheep-cell agglutination tests on normal sera differed considerably from those of Stuart, Burgess, Lawson & Wellman (1934) whose technique was used. The reason for this difference was not determined. It was thought to have some bearing upon the interpretation of sheep-cell agglutination tests in the diagnosis of glandular fever.

3. Although in the glandular fever sera the horse-cell titre was usually higher than the sheep-cell titre, the normal range for horse cells was also higher, and in the diagnosis of glandular fever there did not appear to be any advantage in using horse cells instead of sheep cells.

4. The value of absorption tests in the diagnosis of glandular fever is discussed and a new technique described which has practical advantages over other methods although it does not embody any important new principles.

5. Of 300 normal sera examined by this technique, five contained small amounts of an antibody indistinguishable from glandular fever antibody and ten others contained a sheep-cell agglutinin apparently, though not certainly, different from any yet recognized in human serum.

6. A comparison of the results of the two tests on thirty-one samples of glandular fever serum showed the absorption test to be a more sensitive diagnostic test for glandular fever than the ordinary sheep-cell agglutination test.