Philological studies in Africa are inevitably bound up with the vexed question of the classification of its languages. Many authorities have attempted this classification. Some, like Oust, Drexel, Delafosse, van Bulck, rely on evidence which is as much ethnological as linguistic, while others, like Müller, Meinhof, Westermann, Werner, Sir Harry Johnston, Guthrie, Homburger, Greenberg, and the various authors of the ‘Handbook of African Languages’, have tried to confine themselves to linguistic criteria. It is the purpose of this article to discuss the sort of linguistic data these investigators have consulted, and their methods of using it.