The social impact of mental illness, like that of other diseases which are covered by public health organizations, is governed by quantitative considerations. Two cases of typhoid fever are worse than one, and the same applies to schizophrenia. A principle, complementary to Bentham's doctrine of the greatest good for the greatest number, is the foundation of mental hygiene. The object is to try to reduce the total quantity of discomfort in the community due to mental illness. The effort is not expended only on people recognized to be ill enough to need hospital care. It includes those cases in the community who are less acutely ill and the members of the normal population, who have to tolerate the burden of mental illness in their relatives or associates. The basis of the whole activity is clearly quantitative. In order to comprehend the problem or to evaluate results of effort, adequate statistical data must be collected. It might be thought that to labour this point was unnecessary, and some may even consider it dangerous. Nothing is duller than the compilation of unnecessary statistics, official or unofficial, and nothing can be more misleading than numerical data collected or interpreted without proper forethought. I will therefore first draw attention to some erroneous conclusions, which have been sometimes drawn from statistical data in the psychiatric field.