ments and misstatements which are justified by the desire to "save vaccination from reproach." Thus it happened that till 1881 no deaths were regularly recorded as due to vaccination, although an increasing number of such deaths now appear in the Registrar-General's Reports; while a few medical men, who have personally inquired into these results of vaccination, have found a large amount of mortality directly following the operation, together with a large percentage of subsequent disease, often lasting for years or during life, which, except for such private enquiries, would have remained altogether unknown and unacknowledged (pp. 18–22).
The same desire to do credit to the practice which they believe to be so important leads to such imperfect or erroneous statements as to the vaccinated or un-vaccinated condition of those who die of small-pox as to render all statistics of this kind faulty and erroneous to so serious an extent that they must be altogether rejected. "Whether a person dies of small-pox or of some other illness is a fact that is recorded with tolerable accuracy, because the disease, in fatal cases, is among the most easily recognised. Statistics of "small-pox mortality " may, therefore, be accepted as reliable. But whether the patient is registered as vaccinated or not vaccinated usually depends on the visibility or non-visibility of vaccination-marks, either during the illness or after death, both of which observations are liable to error, while the latter entails a risk of infection which would justifiably lead to its omission. And the admitted practice of many doctors, to give vaccination the benefit of any doubt, entirely vitiates all such statistics, except in those special cases where large bodies of adults are systematically vaccinated or revaccinated. Hence, whenever the results of these imperfect statistics are opposed to those of the official records of small-pox mortality, the former must be rejected. It is an absolute law of evidence, of statistics, and of common sense that when two kinds of evidence contradict each other, that which
