“…On the basis of this writer's studies of the data emanating from collaborative programs originating in the USA (Flanagan, 1969), the United Kingdom (Poole, 1972), France (Roubault et al, 1970;de La Roche and Govindaraju, 1973a, b), the Soviet Union (Kukharenko et al, 1968), Japan (Ando et al, 1971(Ando et al, ,1974, South Africa (Russell et al, 1972) and Canada (Abbey et al, 1975a) the bulk of the evidence points to inter-laboratory factors as the major source of discordant data. That is not to deny the importance of sample homogeneity, sub-sample weight, fineness of grindmg, the need to perform controlled replicate analyses nor the usefulness of statistical data on methods.…”