460 
CONGRESS ON TUBERCULOSIS. 
though even then it might be advisable to take measures to pre¬ 
vent the few cases. But the statistics are not by any means 
unanimous, and those that are likely to appeal with most force 
to the people in this country are not at all in accord with those 
quoted from Germany. During the last few years the evidence 
obtainable from the post-mortem records of two of the largest 
hospitals for children in this country have been analyzed with 
great care, in order to see what evidence they afforded as to the 
relative frequency of the different methods of infection in tuber¬ 
culosis. In the case of the Hospital for Sick Children in Great 
Ormond Street this has been done by Dr. George Still, and in 
the case of the Royal Hospital for Sick Children in Edinburgh 
by Dr. Shennan. The conclusion at which Dr. Still arrived was 
that in 29.1 per cent, of the cases of tuberculosis in children 
primary infection appeared to have taken place through the 
intestine. . That is very far from being an insignificant propor¬ 
tion, and it is a striking fact that Dr. Shennan arrived at an 
almost identical conclusion, and estimated that 28.1 percent, of 
the cases of tubeiculosis among children in Edinburgh are due 
to alimentary infection. There does not appear to be any ground 
for supposing that there is a large margin of error in these sta¬ 
tistics, as the number of cases dealt with was considerable (547 
in the two series), and in both series the post-mortem appearances 
were interpreted in a way to which no exception can be taken. 
In face of these statistics it is not possible- to assent to the state¬ 
ment that cases of primary tuberculosis of the alimentary canal 
are extremely rare. Precisely the contrary conclusion is the one 
that must in the meanwhile be drawn with regard to the state 
of affairs in this country, viz., that, at least in children, primary 
infection by w T ay of the alimentary canal is comparatively 
common. 
I therefore submit that there is still a strong prima facie 
case against animal tuberculosis as a possible source of human 
tuberculosis, and it becomes necessary to consider whether there 
are any data from which one may estimate the extent of the 
danger to which human beings are exposed through the occur¬ 
rence of tubercle bacilli in milk. 
The evidence in favor of the view that the ingestion of 
tuberculous milk is one of the causes of human tuberculosis in¬ 
cludes a number of recorded cases in which the relationship of 
cause and effect appeared to be obvious. From the nature of 
the circumstances, evidence of this kind is very scanty, and it 
must be admitted that very few of the alleged examples are abso- 
