To use a series of statements as they are given would be in the majority of cases quite impossible. All the contradictions imaginable would be present, and we should either helplessly drop any attempt at forming an opinion, or we should get out of the difficulty by a purely arbitrary act. We could by an act of faith believe in some of our writers and accept only what they say or what confirms their opinion, and completely ignore any contradictory information. That would even enable us to form a much more certain and detailed view on many points. Our way of proceeding compels us often to relinquish a very precise, definite opinion, which we could hold if we accepted one statement to be ultimately true, and neglected the others; but it gives us at least the conviction that any more precise conclusion would be unfounded. That all the corrections must be carried out on grounds of ample justification and in the most discreet way is quite clear. It will be seen that in the subsequent pages only rarely have statements been amended, and then the reasons are always given. But it is important that even these few corrections should be done systematically. The above indications will, I trust, help to a certain degree to justify the method adopted in dealing with evidence. Our methodological considerations were necessarily taken on broad lines. To give a detailed and precise description of the method of treating the Australian material would require a whole volume, for there are in all individual cases so many influences and possibilities that may be considered as sources of error, and so many elements to take into consideration, that it would be nearly impossible to trace all the mental processes that have to be followed here. I found it also impossible to give explicitly all my reasons in each place where I ventured to correct a statement. Nevertheless, I have not thought it superfluous to give in outline the chief points adopted in this criticism. In the first place even these general hints will be quite sufficient to indicate the writer's motives to every one who has had to deal in an analogous way with ethnographical materials. And then they will serve as a proof that these questions, doubts, and precautions, were present to his mind while weighing the evidence. In the last place, as science is essentially based on mutual help and mutual agreement, if we had a whole series of workers on a given ethnographic material, a certain general assent, if such could be obtained, would undoubtedly be the best criterion of reliability of sources. But matters should be openly and explicitly discussed.