482 BACILLI IN CHRONIC INFECTIOUS DISEASES. 
In Hirschberger’s investigations milk from tuberculous cows in- 
duced tuberculosis in guinea-pigs, when injected subcutaneously or 
into the peritoneal cavity, in fifty-five per cent of the cases studied 
(twenty). The conclusion is reached that the milk may contain tu- 
bercle bacilli even when the udder of the cow is not involved. Ernst 
also, from an examination of the milk from thirty-six tuberculous 
cows in which the udder was apparently not involved, found the 
tubercle bacillus by microscopical examination in five per cent of the 
samples examined (one hundred and fourteen). 
The prevalence of tuberculosis among cattle is shown by numer- 
ous investigations, and especially by the official inspections of 
slaughtered animals made in Germany. Thus in Saxony, in the 
year 1889, of 611,511 cattle examined 6,135 were found to be tubercu- 
lous (about one per cent); in Berlin, 1887-1888, out of 130,733 ani- 
mals slaughtered 4,300 were found to be tuberculous (3.2 per cent). 
In view of the facts stated the great mortality from tubercular dis- 
eases among children, many of whom are removed from other prob- 
able sources of infection, is not difficult to understand, and the 
practical and simple method of preventing infection in this way, af- 
forded by the sterilization (by heat) of milk used as food for infants, 
must commend itself to all. 
BACILLUS TUBERCULOSIS GALLINARUM. 
The researches of Maffucci (1889) and of Cadiot, Gilbert, and 
Roger (1890) show that the bacillus obtained from spontaneous tu- 
berculosis in chickens, although closely resembling the bacillus of 
human tuberculosis, is not identical with it, varying especially in its 
pathogenic power. This view is sustained by the observations of 
Koch, who says in his address before the Tenth International Medi- 
cal Congress (Berlin, 1890): 
‘The care which it is necessary to exercise in judging of the characters 
which serve to differentiate bacteria, even those which are well known, I 
have learned in the case of the tuberclebacillus This species is so definitely 
characterized by its staining reactions, its growth in pure cultures, and its 
pathogenic qualities, and indeed by each of these characters, that it seems 
impossible to confound it with other species. Nevertheless in this case also 
one should not rely upon a single one of the characters mentioned for de- 
termining the species, but should follow the safe rule that all available 
characters should be considered, and the identity of a certain bacterium 
should only be regarded as demonstrated when it has been shown to corre- 
spond in all of these particulars. When I made my first researches with 
reference to the tubercle bacillus I was controlled by this rule, and tested 
tubercle bacilli from various sources, not only with reference to their stain- 
ing reactions, but also with reference to their growth in culture media and 
pathogenic characters. Only in the tuberculosis of chickens I was not able 
to apply this rule, as at that time it was not possible for me to obtain fresh 
material from which to make pure cultures. As, however, all other forms 
