“…tion between the two is not always possible and, although success has been reported in various eases using this test, it has led to negative results (Whittaker, 1953;Eriks & McGillavry, 1954;Paton & MacDonald, 1957) and even false results in a few cases (e.g., see Robertson & Shearer, 1956;Robertson, Shearer, Sim & Watson, 1958). However, if we modify the original probability distri- to distinguish between the two cases with a large margin of safety, if the test is carried out using the function P(y).…”