beside the lance, club and sling, that he did not enumerate it among the weapons of the people.
The definite statements of the travellers of the eighteenth century, however, can leave no doubt about the more ancient use of the bow and arrow as a weapon and the evidence of Behrens is peculiarly valuable in that he mentions both »Pfeilen» and »Wurf-Pfeilen», showing that he did not confuse arrows with lances either in observation or memory.
The evidence suggests that the bow and arrow was becoming a less important weapon during the interval between the visits of Roggeveen and Duperrey and that this process has continued and led to the total disuse of the bow and arrow in war.
Stephan and Graebner [1] discount the value of the evidence of Behrens and Lesson and suppose that Lesson confused New Ireland with Buka, although the way in which the essential fact is recorded makes this most improbable. More recently Graebner [2] has cited the statements of Behrens and Lesson as examples of untrustworthy evidence and as instances of a mistake liable to be made when dealing with widely distributed objects. The example has not been well chosen; it would be difficult to find an ethnographical fact with better credentials. If the statements of three independent eye-witnesses [3] are not to be trusted, where are we to turn for satisfactory evidence?
W.H.R. Rivers
.
