“…With the mathematical technique of Woolf (1955) it is possible to combine the various authors' series. It will be recalled that a statistically significant excess of group A in gastric cancer was not claimed by all investigators, yet when all the series were combined (Aird et al, 1953;Walther et al, 1956;Wallace, 1954;K0ster et al, 1955;Speiser, 1956;Buckwalter et al, 1956a;Billington, 1956b;Hollander, quoted by Aird et al, 1953) it was shown by Fraser Roberts (1957) that there was an overall association with group A, and that the centres were not significantly heterogeneous. Using the same method, he demonstrated an excess of 0 in patients with gastric ulcers, from series quoted by Aird et al (1954), Clarke et al (1955), K0ster et al (1955), Brown et al (1956), Heist0 (1956), Speiser (1956), Buckwalter et al (1956b.…”