15P> 
R. KOCH. 
theless it is not advisable permanently to keep rabbits and guinea pigs in the same 
rooms that contain tuberculous animals; they would scarcely remain free from 
tuberculosis longer than eight to ten months in infected stalls. In one case a 
number of animals were kept as long as possible by way of experimenting on their 
immunity, but in spite of the best care only here and there one remained free from 
tuberculosis for more than a year, and even these a few months later also became 
victims to the disease. After all these experiences, all the numerous experiments 
in which tuberculosis was conclusively shown, have little or no force as proof 
unless the product itself should make manifest that a spontaneous tuberculosis 
exists or can be excluded. 
As to the second source of mistake, the confusing of non-tuberculous knots 
with genuine tubercles, nothing is simpler than to exclude the same. The genuine 
tubercles are infectious and contain tuberculous bacilli; the false do not. Even 
if one will not admit the diagnostic worth of tuberculous bacilli, one must distin¬ 
guish between infectious and non-infectious knots. Therefore, if by an attempt 
at infection, for example if by the inhalation of any substance, some little grey 
knots be caused in the lung of a dog, one may not content himself with this simple 
result, and resting upon this claim that these are genuine tubercles. Under all 
circumstances the infectious nature of such little knots must be proved. Where 
genuine tuberculosis is concerned one is generally spared the trouble of especially 
proving their infectious nature by inoculation with the knots, for in this case the dis¬ 
ease seldom shows itself confined to one spot; almost always it has already attacked 
other organs of the body, itself furnishing proof of its infectious nature by its prop¬ 
agating ability. Therefore, when the formation of tubercles stretches itself out 
past the original infectious spot into the lymph glands, lungs, liver and spleen, it can 
without further proof be considered infectious. If, as is, for example, the case after 
the inhalation of non-virulent firm particles into the lungs, and after the injection 
of granular masses into the “bauchhohle ” (belly cavity) in the peritoneum, the knots 
caused by this remained confined to the place of infection (here the lungs and 
peritoneum) and show no inclination to further infection of the body, then this 
circumstance speaks against the inference that genuine tubercles exist here, and 
special proof of their infectious nature must be furnished. If this is not done, as 
it, in an incomprehensible manner, has not been in several of the newer investiga¬ 
tions undertaken to prove the non-infectious nature of tuberculosis, then the real 
proof is lacking in the experiment. 
The third mistake mentioned, the unintentional infection by means of instru¬ 
ments, etc., appears to have clung to almost all former investigations with regard 
to tuberculosis to a greater or less degree, whether such investigations were directed 
towards proving or disproving its infectious nature. And yet this mistake may 
be avoided without great difficulty if one holds to the rules concerning anti¬ 
septic operations, and above all things carefully disinfects the instrument for 
every single attempt. All metal instruments, such as scissors, pincers, knives, 
inoculating lancets, must be heated thoroughly. Special care is demanded in 
the treatment of the syringes used for injection. Syringes, of ordinary con¬ 
struction cannot be disinfected with sufficient certainty, because the}' do not 
admit a high degree of heat without being injured, and liquid means of dis_ 
infection, as shown by experience, do not certainly destroy the infectious 
