“…Where only the denominator of the ratios had to be estimated, the numerators being "raw" frequencies (see Early Fetal Mortality, p. 1680), the estimators would make too little allowance for any abrupt, shortlived changes in the numbers at risk, hence the total amount of variation in the monthly ratios would be too large. Ingalls, 1953) have suggested that a community with large Jewish representation might be expected to have a relatively low incidence of CNS malformations, and that the frequency of anencephalus in particular has been reported to show wide geographical variation (Penrose, 1957 (Coffey and Jessup, 1959) suggested that an attack by this virus during pregnancy could double the risk of fetal OCTOBER, 1962 malformations as a whole. If the same risk factor* had applied to the New York victims of this epidemic, some associated excess of malformations would have been likely to show up in the present analysis, provided that about a quarter or more of the women pregnant during the epidemic were attacked.…”