ETHNOLOGY HEDLEY. 



255 



tempted to believe that the German traveller had before him a 

 Rotatory Adze, though the distinguishing feature of it escaped 

 his observation. My reasons for this opinion are that the shell 

 blade is shown not directly connected with the handle, but in- 

 serted into a separate holder which is in turn fastened to the 

 handle ; and further that in the immovable adzes the method, 

 which I have already described, of lashing the blade to the 

 handle, is quite different, whereas the mode and lashing of the 

 Caroline adze is exactly that of the Pelew Rotatory Adze, namely 

 one series of backwardly and another of forwardly directed cords, 

 arising from opposite sides of the handle and meeting above. 

 This arrangement is seen again in an axe-adze Finsch figures 

 from Guap, near D'Urville Island, German New Guinea.* The 

 drawings of Edge-Partington are not sufficiently elaborated to 

 permit much appeal to detail, but the points just discussed 

 suggest to me that an adze, figured as from Pitcairn Island,! is 

 probably a Rotatory Adze. Recollecting that the " Bounty " 

 mutineers found Pitcairn uninhabited, I regard this locality with 

 suspicion. Others figured as from the Carolines, Santa Cruz, 

 New Guinea and New Zealand (!)J may perhaps belong to the 

 group under consideration, as may that shown on p. 313 of 

 Codrington's Melanesians. 



If it be accepted, as it generally is, 

 that the Plane-iron Adze is the direct 

 descendant of the Stone or Shell Adze, 

 then it cannot be denied that the Rota- 

 tory Adze 1 here figure is derived by 

 parallel descent from an adze like that 

 figured by Keate. Various aspects of a 

 specimen of the Rotatory Adze now 

 in common use in Funafuti, where it is 

 called " atupa," are shown by fig. 19. 

 The handle of the atupa differs from 

 that of the toki, in that the short arm 

 is produced so as to transform the 7 into 

 an oblique and unsym metrical T. The 

 example selected for illustration weighs 

 one pound, six ounces ; the handle is two 

 feet long and the head half as much. In 

 this particular instance the cutting edge 

 is a European hoe-blade; in another, part 

 of an iron door-hinge has served, and 

 probably scrap-iron in almost 'any form 





Fig. 19. 



* Finsch Efchnol. Atlas, pi. 1, fig. 7. 

 t Edge-Partington loc. cit , ii., pi. xv., fig. 5. 



J Loc. cit, ii., pi. xciii., fig. 3 ; i., pi. ccc., fig. 3; pi. ccclxxx., fig. 3; 

 pi. clxii., fig. 4. 



