[Analysis of the geographical variations of twinning rate in France. An indirect approach to study spontaneous abortions (author's transl)].

The hypothesis of the existence of spatial variations of spontaneous abortion rates has two logical consequences on dizygotic twinning rates: they must be heterogeneously distributed and show a positive correlation with natality. French data available on dizygotic twinning rates from 1950 to 1974 are in agreement with both implications. Although alternative explanations are possible, these findings do reinforce the relevance of dizygotic twinning rate variations as a tool for the assessment of reproduction failures.