I know full well that when I speak of the Germans of Tacitus or of the Aryans of the Veda as the _ci-près_ of primitive man, all the indications of modern, or at all events of secondary and tertiary thought which I have pointed out myself in the hymns of the Rig-Veda, and which might easily be collected from the book of Tacitus, will be mustered against me. Must I quote the old saying again: _Est quoddam prodire tenus si non datur ultra_? All I maintain is that these historical documents bring us as near to the primitive man as historical documents can bring us; but that the nearest point within our reach is still very far from the cradle of the human race, no one has pointed out more often than myself. There is, however, plenty of work still to be done in slowly following up the course of human progress and tracing it back to its earliest stages, as far as literary, monumental, and traditional documents will allow us to do so. There are many intricate windings of that historical river to be explored, many riddles to be solved, many lessons to be learnt. One thing only is quite certain—namely, that the private diary of the first man will never be discovered, least of all at Cape Horn. I have thus tried to show how untenable is the theory which would boldly identify the modern savage with primitive man, and how cautious we ought to be whenever we take even a few hints here and there from degraded tribes of the present day in order to fill out our imaginary picture of the earliest civilisation of our race. Some lessons, and even important lessons, may be learnt from savages, if only they are studied in a truly scholarlike spirit, as they have been, for instance, by Callaway and Codrington, by Waitz and Tylor. But if the interpretation of an Homeric custom or myth requires care, that of African or Polynesian customs or myths requires ten times greater care, and if a man shrinks from writing on the Veda because he does not know Sanskrit, he should tremble whenever he writes the names of Zulus, unless he has some idea of what Bântu grammar means. In arguing so far, I have carefully kept to the historical point of view, though I am well aware that the principal traits in the imaginary picture of primitive man are generally taken from a very different source. We are so made that for everything that comes before us we have to postulate a cause and a beginning. We therefore postulate a cause and a beginning for man. The ethnologist is not concerned with the first cause of man, but he cannot resist the craving of his mind to know at least the beginning of man.