BY C. HEDLEY. 291 



On dissecting the waist band of a New Guinea dress, I find 

 another knot so different and so compHcated that it also seems 

 worthy of publication."^ Other knots have since come under my 

 notice, and I can commend the subject to students as likely to 

 repay careful investigation. A trifle like one of these knots 

 might serve to trace migrations or affinities, for these would 

 exclusively descend from woman to woman — there as elsewhere 

 the most conservative element in the population. Every collec- 

 tion is well supplied with material, and these dresses are among 

 the last of native fabrics to be obliterated by European civilisation. 



The dress containing the knot to be described is a kind usual 

 in East British New Guinea, dyed in alternate vertical stripes, 

 with a scalloped flounce at a quarter depth. Inside (fig. 3) the 

 belt shows the fibres as if in two beaded rows, outside (tig. 4) the 

 fibres appear caught in a chain stitch. A glance at the exterior 

 gives no idea of the intricate knot shown unravelled (fig. 5). Two 

 bundles of fibres are disposed in a series of three. The waist 

 band of the titi is always of two strings. Over the lower is 

 hitched the two fibre bundles; then dividing they receive between 

 them the fibres of the preceding series; closing again, they divide 

 the succeeding pair; then passing over the top cord, they descend 

 behind the second and third pairs, and opening out contribute to 

 the skirt of the dress. 



EXPLANATION OF PLATES. 



Plate XIV. 



Fi^s. 1 and 2.— Palu hook from front and side. 



Plate XV. 



Fi2. 3.— Inside belt of titi. 

 Fig. 4. — Outside of same. 

 Fig. 5. — Knot unravelled. 



