More complicated forms of the multifactorial hypo-thesis, such as one involving a variation of expressivity of the genes with age, would require a more complex scoring system. The need for tnis would be indicated, if sufficient data were available, by the appearance of particular patterns of heterogeneity in the scores. Some forms of the single-gene hypothesis-for example, one in which the same fraction ofthe upper end of the distribution of pressures was hypertensive, whatever the actual pressure or the age of the subjects-wouldalso be satisfactorily explored bythe use of age-adjusted scores. Other forms would not-for example, any in which a particular pressure was regarded as the boundary of normality. In this case segregation into classes with andwithout hypertension couldbe obliterated by the pooling of age-groups in which a varying proportion of subjects attained the critical pressure.

We make no apology for basing our exploration of the data on the multifactorial hypothesisin its simplest form. In the absence of an alternative hypothesis of sufficiently detailed specification it is the natural hypo-thesis to use with data appearing as continuously distributed measurements, and if it is wrong we believe this will be revealed by internal contradictions in the data.