KBCOHDS OF THE AUSTRALIAN MUSEUM. 



prevent them from slipping ; over the wider portion of the raft^is 

 placed a bundle of dried giass, upon which the traveller squats 

 and paddles himself along. 



7. Bark-Canoes are made either of one. two or three sheets of 

 bark, and in the main are river-craft, though on the East Coast 

 they are often taken across to the neighbouring islands, and on 

 the West Coast out on the sea, but only wlien the wind and 

 weatlier are favourable. 



Those liuilt of a single sheet are found on the Gulf Coast, ex- 

 tending from ihe Batavia and Ducie Rivers down to the Archer 

 River, and on the eastern littoral along an area reaching from 

 the Johnstone Kiver to a little below Cardwell. That their area 

 of distribution on the latter coast was much further south than 

 this within very recent times is rendered highly probable from 

 the fact that the Keppel Islanders, who possessed no canoes wiien 

 I first came amongst them, made me models of the single-sheet 

 type to explain tlie craft they u.sed to have in days gone by. 

 These models were all the more interesting in that the only 

 traces of l)ark canoes that were discoverable amongst the 

 neighbouring mainland natives of the Fitzroy River were of the 

 three-sheet type. 



Tiie manufacture of such single-sheet canoes is practically the 

 same on both coast-lines, the existing differences being only in 

 detail. At tlie TuUy River (East Coast) the bark employed is 

 obtained from at least five different timbers, known under their 

 local IMallanpara names as nupa, kirau, kiri, yabandai, and 

 kalkara, of which only the first has been identified as Calophi/l/um 

 tomentosum, Wight. The method of stripping has alread}' 

 been explained"^. I was further informed that the bark 

 from these particular trees will strip more or less at any 

 time of the year, i.e., not necessarily onl}' at the end of 

 the wet .season when the sap is up. The sheet of bark, 

 according to length required, having been removed, one 

 of its ends is heated over a tire to render it pliable, and 

 thewhole length then folded long ways, with the outerside 

 of the bark outwards ; the end which has been heated is 

 next clamped in a vice. This vice is made of two 

 switches (tig. 3) tied tightly below around a stiff bundle 

 of grass, bark, etc., so as to form a kind of fork, the ' leg ' 

 ofwhich is implanted firmly into the ground ; the 'arms* 

 are sul)sequently tit-d ov(>r the eiul of the folded length, 

 which is thus held tightly in position (fig. 4). The name 

 Fig. 8. given to this piece of apparatus is yuku nambil-nainbil 

 (yuku:=:tree, log, tiuil)er ; nambil:=to squeeze). A spreader or 



3 Roth— Bull. 7— Sect. 1. 



