The article as a whole argues that the observations of Leo Frobenius in West Africa between 1907 and 1912, though they are commonly ignored by Anglo-American archaeologists and anthropologists, contain material which is of value to the historian of early twentieth-century West Africa. The author examines Frobenius's methods and observations in the light of his own statements (some of them in works which have not been translated into English or French) and suggests various outside sources which could be used to check the reliability of Frobenius's accounts.
The attitude of Anglo-American anthropologists is explained in terms of developments in anthropological theory since the First World War.
The article examines the defects of Frobenius's anthropological method, as exemplified by his work among the Kabré of North Togo shortly after their ‘pacification’ by the Germans; and suggests that the observations themselves may be more reliable than Frobenius's method would lead one to expect.
The article then considers Frobenius's archaeological method as exemplified by his work in Ife, argues that he was somewhat unscrupulous in forcing the sale of artefacts, and also, that by failing to keep adequate site records or even to supervise the digging himself, he destroyed a considerable amount of archaeological evidence at Ife. Historians should, however, study his work in Ife to note where such destruction has taken place, and what kind of evidence has been destroyed.
The article suggests outside sources, both oral and written, which might enable historians to check the veracity of Frobenius's account of his own method, and also the reliability of the observations themselves. Finally, it is suggested that an assessment of the reliability of his observations in Togo and Nigeria might be generally indicative of the reliability of his observation throughout West Africa in the period 1907–12.