View report page

120
Annual Report of the London County Council, 1910
lactose-fermenting) content of the milk becoming greater in the hot weather as might be expected,
but the mortality curve does not exhibit any correlation. For instance, the first week of July, 1910,
shows a fairly high average temperature, and a very high percentage of "fermenters," and yet the
mortality is below the average of the whole year and much below the average of June or August of
the same summer.
Chart II. gives the mean monthly temperature and the percentage mortality, and here again
no close relationship is apparent between the mortality curve and the temperature.
Results of subcutaneous injection of "single cow" milk samples.—To make this section clear,
it is necessary to explain shortly the method of procedure, after a milk from a churn has been proved
to contain tubercle bacilli by producing tuberculosis in a guinea-pig. The herd from which the milk
in the corresponding churn comes is inspected by a veterinary expert, and any cow showing evidence
of udder disease in any form, or of general illness, is carefully examined and its milk is forwarded for
examination for the presence of tubercle bacilli. The total number of such samples in the period dealt
with in this investigation was 193, of which 69 (— 35.75 per cent.) were proved to contain B. tuberculosis.
These single cow milks," coming thus from diseased or suspected cows, frequently contain
pus and organisms associated with the disease, and it was of interest to enquire whether the milk from
these cows produced a higher mortality than the ordinary "mixed milks" obtained from many cows
the majority of which are presumably healthy. A comparison of the figures does not, however, show
any great difference. The number of cases where both animals died, and a repeat specimen was
necessary, is here 14 in 193 samples (— 7.2 per cent.), as compared with 736 in 4,508 (= 16.3 per cent.)
while the total "non-tubercular mortality" is here 24 per cent. and in the general samples 33 per cent.
A comparison of the period lived after inoculation by the two groups does not reveal any notable
differences.
Consideration of cases in which both animals injected with the sediment of the same sample died
—If the substances which cause the "non-tubercular mortality" were in the milk, and if they were
the only factors present, one would expect the greater number of milk samples to cause the death of
both animals, or of none, since each received an equal amount of the emulsified sediment. Yet this
is not what we find, for (see table on page 115) in the first 4,508 samples of milk there were 2,333
(=736 + 1,597) instances, in which one or both of the injected guinea-pigs died, and there were only
736, i.e., less than one-third, in which both animals died.
A consideration of those cases in which both the animals died did not suggest any ready explanation
of this result, but suggested, on the contrary, that the lethal agent was in the milk. In analysing
them, we find that the majority died within 24 hours of each other ; e.g., in the last 72 instances, 52
of the pairs of animals died within 24 hours of each other. The majority of such "double deaths"
occurred prior to the period of the present investigation, and of those recorded during the past year
the greater number occurred during the "epizootic period" in November and December, 1909, the
bacilli found in one or both of the dead animals being generally the B. aertryck, or, less frequently,
the B. pseudo-tuberculosis rodentium. At this time the presence of a filter-passer" amongst both
the stock and the inoculated animals was indicated as the probable cause of death.
In some of the cases occurring since that time the plates inoculated from the organs of one or
both of the dead animals remained sterile, but in a number of others the same bacillus was obtained
from both animals at autopsy, the organism isolated being identical with, or closely allied to, the
lactose fermenters common to MacConkey's table of faecal organisms and to Orr's table of milk
organisms. We have seen that many of these organisms are lethal to guinea-pigs, when injected subcutaneously,
so that the facts brought forward above raise a strong presumption that the infecting agent
was often injected with the milk.
Discussion of results.
Since the intital problem was to ascertain the actual causes of the non-tubercular mortality
the most obvious method of doing this was to see if the organisms isolated from the tissues of the dead
animals were pathogenic to guinea-pigs when injected subcutaneously. If these organisms were found
in the milk ejected, or had been shown by other observers to be commonly present in ordinary London
milk, strong presumptive evidence would be afforded that such organisms were the real causes of the
mortality.
But difficulties arising out of this train of reasoning early presented themselves. From many
of the animals dying during the early part of the investigation the B. aertryck was recovered, yet there
were the strongest reasons for believing that this bacillus had not been injected with the milk sediment.
Similarly the B. pseudo-tuberculosis rodentium was isolated from the dead animals, sometimes from
the two animals of a pair dying on the same day, yet here again the evidence available strongly suggested
that this bacillus was not derived from the milk.
When dealing with the lactose fermenting organisms isolated, the chain of events at first seemed
clearer. The bacilli isolated from the dead animals were faecal organisms of types found frequently
in milk and were actually obtained from some of the samples of milk injected ; they were, moreover,
lethal to guinea-pigs when injected subcutaneously.
One must not, however, conclude off hand that the presence in the milk sediments of organisms
originally derived from faecal pollution was entirely responsible for the deaths. Many of these
organisms have also been found in the intestine of the healthy guinea-pig. In the case of the guineapigs
inoculated with milk they may have been so derived, as it is not uncommon for these intestinal