92 TYPHOID FEVER IH DISTRICT OF COLUMBIA. 
weeks and at best of many months. At the same time the other 
effective conditions are also heightened. It is no wonder, then, that 
our experiments show a reduction of over 99 per cent typhoid 
bacilli frozen, and we may be sure that in nature the destruction 
would exceed rather* than fall short of such a limit. 
This reduction obtains in tubes which are frozen solid, where there 
is no chance for mechanical extrusion. In natural ice there is another 
purifying influence. Of the germs remaining in the water at the 
time of freezing, 90 per cent are thrown out by the physical phenom- 
ena of that process. This reduction is separate from and supple- 
mentary to the disinfecting action of the cold. Accordingly, when 
both factors work together it is obvious that only one out of a thou- 
sand t}^phoid germs present in a polluted stream has a chance of 
surviving in the ice. 
Under natural conditions the pathogenic germs present in the 
most highly polluted stream are comparatively few. Of these few 
one-tenth of 1 per cent may be present in ice derived therefrom. But 
even these scattered individuals are weakened by their sojourn under 
unfavorable conditions, so that, as we have seen, they require nearly 
twice as long for their development as do the normal germs, and 
these few and weakened germs very likely could not produce many, 
if any, cases of typhoid fever, for vitality and virulence in disease 
germs are probably closely related. 
^AYith artificial ice the case is somewhat different, for such ice is 
made from water frozen solid and is as a rule quickly consumed. 
Artificial ice, if made from pure water, should be above reproach; 
but if it be made from water that is impure it may contain the germs 
of infectious disease; and inasmuch as artificial ice is used quickly 
after its manufacture, the possibility of purification by time is 
excluded, and such ice might therefore conceivably be a menace to 
the public health. 
^AYith natural ice, as long as absolute sterilization is not effected, 
there must always remain a certain element of doubt, as in the case 
of sand filters alluded to above or in the practice of room disinfection 
after contagious diseases. The thickness of a layer of ice is often 
artificially increased b}^ cutting holes in it and flooding that already 
formed with the water of the pond. In such a case the effects of 
crystallization are excluded, as in the laboratory tubes. Ice thus 
formed might be cut at once and served within a week or two, and in 
such an exceptional case we can not say that sufficient of the virus 
might not persist to excite the malady. Yet such an instance must 
be very exceptional, and the general result of human experience, the 
absence of epidemics of typhoid fever traced conclusively to ice, the 
fact that cities like New York, and Lowell and Lawrence in Massa- 
chusetts, have used the ice of polluted streams and have yet main- 
