ORIGIN OF EPIDEMICS O47 
the evidence is complete—when and how that route became infected, 
when and by what sub-routes the infection was distributed, why it 
infected the patients found and not others, etc. 
“In this illustration I have assumed complete ignorance on the 
part of the epidemiologist as to everything connected with the com- 
munity he is investigating, except what he finds by cross-examining 
the patients. As a matter of fact, every epidemiologist, however much 
a stranger to the particular community he enters, begins to learn about 
it from the moment he enters it. 
‘“‘ Thus almost unconsciously he notes the size of the town and com- 
pares it with the number of cases reported as existing; if it is summer 
time he almost automatically notes the presence or absence of open 
toilets in the back yards, of manure piles and of garbage cans—all 
bearing upon fly infection. If it is winter time or the community be 
well sewered, he does not even consider flies. If the cases are grouped 
in one quarter of the town, while the public water supply extends all 
over it, he tentatively eliminates the water supply, before he asks a 
question. If good surface drainage and a sandy soil exist, or driven 
wells are chiefly in vogue, he tentatively eliminates well water—even 
before he registers at the hotel. 
‘“* This is not and cannot be a complete synopsis of all the combina- 
tions of circumstances which the epidemiologist meets. It is intended 
to illustrate his methods and to show why they are incredibly rapid 
and incredibly accurate—how they eliminate speculation and guarantee 
a correct solution—which means of course the achievement of the 
great end, the finding of proper measures for suppression. 
‘¢ As soon as the route is indicated, he must go to that route, and 
establish beyond peradventure that it was in truth responsible. A 
water supply cannot convey typhoid if typhoid fever discharges have 
not entered it. There is no object in attributing an outbreak to fly 
infection from toilets into which typhoid feces have not been dis- 
charged at such a time as to account for the cases. A milk supply, 
not handled at some point by an infected person, nor adulterated at 
some time with infected extraneous matter cannot convey typhoid. 
Whatever his results, they cannot be true unless they are consistent— 
they should not be accepted unless they are provable—and proved. 
“Tf the public health detective is familiar with the community 
where the outbreak occurs, including its water supplies, its milk sup- 
plies, the sociological relationships of its people, etc., etc., he can often 
tentatively determine the cause of the outbreak by a mere inspection 
of the names and addresses of primary cases, especially if plotted on 
