[...] a good interpretation makes sense of such an impressive amount of mutually corroborating data that coincidence must be excluded. Such a result cannot be reached without sound practice and good method.