Provided the assumptions underlying Weinberg's method of differentiation of types of twin pairs are correct, this analysis of the statistics of twin births in Italy in the two years 1949 and 1950 shows that there are variations in the frequency of monovular twin births greater than can be attributed to chance deviations from a constant frequency of monovular twinning. There is a tendency for the frequency to increase with maternal age, and for young mothers, a tendency for it to increase also with parity. In general, the occurrence of monovular twinning is probably more closely related to maternal age than to parity.
The frequency of binovular twinning appears to be independently related to both maternal age and parity. Amongst births of each rank the frequency of twin births increases to a maximum at maternal ages 35-39 years; and, at each age level, the frequency does not change between birth ranks 1 and 2 but subsequently increases fairly continuously. There is a suggestion that there may be a modal birth rank as well as a modal maternal age, and that binovular twinning may be no more closely related to maternal age than to parity.