1. Data are presented relative to twin pregnancies within the total, the “white” and the “colored” populations of the United States for the 15 year interval from 1922 to 1936, inclusive.
2. Attention is called to the fact that twin pregnancy frequencies in all human populations fall within rather narrow limits, and it is suggested that this fact supports the conclusion based on other lines of evidence that human twinning has some genetic basis.
3. It is shown that the twin pregnancy frequency for the “colored” population is significantly higher than is the corresponding percentage for the “white”. It is concluded that this difference is due, in part at least, to a difference in heredity.
4. It ise stimated that 33.43 percent of all twin pregnancies within the total population are of the monozygotic type and that within the “white” and the “colored” populations considered separately the corresponding percentages are 34.16 and 28.91, respectively.
5. Calculations show that the higher percentage of twin pregnancy frequency within the “colored” population than within the “white” is exclusively the result of a higher percentage of the dizygotic type of pregnancy.
6. It is shown that livebirths among twins have a significantly lower percentage of males than do livebirths among members of single pregnancies. This is attributed primarily to a higher percentage of male deaths in twin than in single pregnancies during early uterine stages.
7. It is shown that the percentage of stillbirth for twin pregnancies is twice as high as is the percentage for members of single pregnancies. Even so it is suggested that there may be at present or in the future a trend started in the direction of an increase in the percentage of plural births within the human species.
8. It is shown that the percentage of sets of twins in which both members are stillborn is higher than the one expected on the assumption that the deaths of the two members are independent events.
9. The percentage of sets of twins in which both members are stillborn is found to be significantly higher for same sex sets than for opposite sex sets. This is attributed primarily to the inclusion of monozygotic sets among the former.
10. It is also shown that the percentage of sets of twins in which only one member is stillborn is significantly higher for same sex sets than for opposite sex sets. This is attributed primarily to the inclusion among the former of monozygotic sets in which one member is handicapped as a result of an unequal twinning process or of some related phenomenon.
11. Many additional differences relative to twin pregnacies are shown in the accompanying tables, but are not discussed either because they are relatively insignificant or because they require supporting data.