The first question to be answered is: what is alcoholism? The only realistic answer to this question is that alcoholism is what it is defined to be. The delimitation of the concept is arbitrary. First of all there are in this respect great differences of opinion between different population-groups. What, in one group, may be regarded as just an innocent or even healthy habit would in a second population-group be stamped as a deplorable moral defect with disastrous consequences, and in a third group just as a pure medical problem. It is true, of course, that those members of the medical profession who are especially interested in research on alcoholism will, irrespective of the part of the world from which they come, have a tendency to converge towards a tolerably homogeneous concept of alcoholism. But even if this is the case the selection of alcoholics which come into their scope will to a high degree depend on the attitude of the population and not on the concepts of the doctor. Such selection factors must more or less unconsciously colour the view of the individual research worker. We know that not only do quantitative and qualitative properties of alcoholism differ very much from one population-group to another, but, in addition, the fraction of alcoholics which comes to medical attention is certainly not representative of alcoholism in general in that population-group. Most materials of alcoholics which have been under research must be very heterogenous, and one must necessarily differ from the other. How could it be possible to compare the genetical back-ground of such quite different and heterogenous materials?