Science
would never make progress under such conditions. Neither the long
incubation of hydrophobia, nor the protecting power of vaccination, would
ever have been admitted, if the results of observation in these
affections had been rejected as contradictory to the previously
ascertained laws of contagion.

V. The disease in question is not a common one; producing, on the
average, about three deaths in a thousand births, according to the
English Registration returns which I have examined.

VI. When an unusually large number of cases of this disease occur about
the same time, it is inferred, therefore, that there exists some special
cause for this increased frequency. If the disease prevails extensively
over a wide region of country, it is attributed without dispute to an
epidemic influence. If it prevails in a single locality, as in a
hospital, and not elsewhere, this is considered proof that some local
cause is there active in its production.

VII. When a large number of cases of this disease occur in rapid
succession, in one individual's ordinary practice, and few or none
elsewhere, these cases appearing in scattered localities, in patients of
the same average condition as those who escape under the care of others,
there is the same reason for connecting the cause of the disease with the
person in this instance, as with the place in that last mentioned.

VIII. Many series of cases, answering to these conditions, are given in
this Essay, and many others will be referred to which have occurred since
it was written.

IX. The alleged results of observation may be set aside; first, because
the so-called facts are in their own nature equivocal; secondly, because
they stand on insufficient authority; thirdly, because they are not
sufficiently numerous. But, in this case, the disease is one of striking
and well-marked character; the witnesses are experts, interested in
denying and disbelieving the facts; the number of consecutive cases in
many instances frightful, and the number of series of cases such that I
have no room for many of them except by mere reference.

X. These results of observation, being admitted, may, we will suppose,
be interpreted in different methods. Thus the coincidences may be
considered the effect of chance. I have had the chances calculated by a
competent person, that a given practitioner, A., shall have sixteen fatal
cases in a month, on the following data: A. to average attendance upon
two hu