The origin of human triploidy: an integration of qualitative and quantitative evidence

SUMMARY 1. The origins of human triploidy are inferred from marker chromosomes by use of maximum‐ likelihood models applicable to all mammals with a simple XX: X Y sex determination mechanism. The numerical data comprise a compilation from the literature of 323 triploids and two 2 N /3 N mosaics, nearly all spontaneously aborted. 2. In the particular data, at the time of conception, it is estimated that digynic triploidy (two sets of maternal chromosomes) is less common than diandric (two paternal sets), and that dispermy (fertilization by two spermatozoa) exists and is commoner than diplospermy (fertilization by a diploid spermatozoon) as a cause of diandry: the best estimate of the diplospermic contribution is zero. A low viability of XYY conceptuses has been detected and measured, and is an essential factor in all interpretations: one consequence is that ˜ 15 % digynic triploidy at conception would change to ˜ 19 % at the time of scoring conceptuses. The viabilities of XXX and XXY are comparable with one another. 3. Inferences about the general population of triploids at conception are based on approximate confidence limits for the estimates. Digyny probably accounts for at most ˜ 41 % of all triploidy (perhaps as little as ˜ 7%). Diandric triploidy forms a major class (˜ 59‐˜ 93%) attributable at least in part to dispermy. The best estimate of diplospermic triploidy is zero, but it could account for up to ˜ 20 % of all triploids and up to ˜ 31 % of all diandric triploids. The low viability of XYY triploids remains marked as a population characteristic. 4. Certain items of indicative ‘ancillary evidence’ are used minimally, primarily as independent checks of the analytical results. They are: (a) that XYY have a low viability, ( b ) XXX and XXY have a comparable viability, ( c ) diplospermy plays little or no part in the genesis of triploidy, ( d ) some triploids are dispermic, ( e ) F ‐body studies can estimate the relative numbers of XX, XY and YY diploid spermatozoa, ( a ), ( b ) and ( d ) are predicted correctly when ( c ) is used to guide analysis: ( a ), ( c ) and ( d ) when ( b ) and ( e ) are so used. This gives some confidence in the validity of the analyses, and in the reality of phenomena established from both qualitative and quantitative evidence. 5. The bulk of human triploidy probably arises near the time of fertilization.