1. This report summarizes the results of investigating 6963 cases of pulmonary tuberculosis in Great Britain.

2. The tubercle bacilli in the sputum of each case were obtained in culture and their types determined.

In Scotland out of 2769 cases 2609 yielded strains of human type (2564 eugonic and 45 dysgonic) and 160 (5·8 %) yielded strains of bovine type.

In England tubercle bacilli of the human type were demonstrated in 3592 cases and of the bovine type in seventy-nine cases. Dysgonic human strains were found in seven cases, four of which occurred in the only series of English cases, namely 680, which were systematically examined for strains of this variety; dysgonic human strains were therefore proportionately less frequent in England than in Scotland. Of the seventy-nine bovine cases fifty-four occurred among 3422 unselected cases and twenty-five among a series of selected cases.

In Wales 203 cases were examined and two were found to be bovine infections.

In Eire no bovine infections were found in a series of 320 cases.

3. The total number of cases of pulmonary tuberculosis shown to be expectorating bacilli of the bovine type of the sputum was 241, but twenty-five of them, occurring as they did among selected cases, are not used in the following percentages.

The proportional frequencies of bovine infections were higher in all regions of Scotland than in England, the percentage being highest in the Orkney Islands (25·8%). The rural districts of the mainland of north-east Scotland follow with 9·1% and then those of the rest of Scotland with 5·2%. The City of Aberdeen gave 4·4% of bovine infections, but many of these had been infected in the country. In England the highest percentages were recorded in the north and middle regions, namely 2·0%, the southern part yielding only 0·6%.

4. The strains from 232 of the 241 cases were fully virulent and from nine they showed varying degrees of attenuation.

5. In six cases the bovine bacilli were associated with tubercle bacilli of another type, five times with eugonic human strains and once with a strain which could not be cultivated.

6. The anatomical evidence (previous cervical and abdominal glandular and bone and joint tuberculosis) in about a third of the cases in Scotland and in a quarter of those in England was strongly in favour of the digestive tract as the channel of entry of the bacilli.

7. Autopsies have been made on fourteen cases. In one case the lungs only were examined. In nine autopsies the anatomical evidence indicated the alimentary canal as the route of infection. In four autopsies the anatomical evidence was inconclusive.

8. A history of tuberculosis was obtained in seven families in each of which two cases of pulmonary tuberculosis occurred. But bacteriological investigations in each of two families showed human in one affected person and bovine tubercle bacilli in the other and therefore disproved human to human infection. All the ten patients in five families yielded cultures of bovine tubercle bacilli. We concluded from the evidence that in one family both cases were of alimentary origin. Human to human infection was presumptive in the remaining four families. No autopsies were made in the last cases.

9. Twenty-five patients were associated in their employment with cattle. Autopsies were made on two of them but the anatomical evidence as to the channel of entry of the bacilli was inconclusive.

10. Of the 241 persons, forty-eight were known to be married and had 120 children. Bovine strains were obtained from two children (two families). Bacteriological evidence disproved infection from the parents in one case but was in favour of it having taken place in the other.

11. One probable instance of infection with bovine bacilli spreading from man to cattle is quoted.

12. A case of tuberculosis of the lungs due to bovine tubercle bacilli is indistinguishable clinically, radiologically and by post-mortem examination from one due to human tubercle bacilli.