REVIEW OF UNDULANT FEVER                       33

agent from market samples of butter, and the various processes in the manufacture
of cheese may have a like effect.

      The number of samples of market butter and cheese examined by Carpenter
and Boak was not very large, but the technique appears to have been sound and
the results conclusive. Up to the present time, therefore, butter and cheese have
not been proved to be likely sources of infection.

      There remain, then, only milk and fresh cream as probable sources of infection,
and proper pasteurisation would appear to be sufficient to render these non-
infective.

      It has been stated above that various workers, when searching especially for
Br. abortus infection in milk, have found upwards of 5 per cent. of samples infected.
If these findings are true, it appears very strange that the presence of the infection
has not been discovered much more frequently and at an earlier date. During the
past 30 years there has been an increasing number of milk samples examined by
guinea-pig inoculation for the presence of the tubercle bacillus. If Br. abortus is so
frequently present in milk samples as the literature on the subject would lead one to
believe, surely it should have been detected in these routine examinations for the
presence of tubercle bacillus, especially in view of the small number of abortus
organisms which are capable of producing the disease in the guinea-pigs. The
author has failed to find any reports of such infection cropping up in routine tests
for tuberculosis. Again, Br. abortus produces in guinea-pigs lesions which have
some resemblance to tubercle, and such lesions, if present, would naturally have
been suspected and subjected to careful examination.

      There would seem then to be a risk of abortus infection from ingestion of cows'
milk, but in view of the relatively small number of cases reported and of the
failure to find the organism in milk unless specially sought for, one is inclined to
the view that the risk is not great. Since, however, it is undoubtedly present,
efforts should be made to exclude infected animals from producing herds, and the
practice of immunising cattle with living abortus cultures should be discontinued,
since such animals have been shown to excrete living abortus organisms in their
milk for a considerable period after inoculation.

                                                REFERENCES.

      (1)  Bruce, D. (1887). Practitioner, London, 39, 161, 188 ; 40, 241.
            „ (1889). Brit. Med. J. 1, 1101.

      (2)  Wright, A. E., and Smith, F. (1897). Lancet,  1,656.

      (3)  Reports of the Commission for the Investigation of Malta Fever, 1-7, London, 1905-07.

      (4) The Oxford Medicine,4,800.

                                                                                          C 2