LAUFER, THE DECORATIVE ART OK THE AMUR TRIBES. 57 



is added to the facing spirals filling the hollows of the wave-line that open down 

 ward, while in those opening upward appears a doubling of the same figure as 

 that in the upper row ; that is to say, from the middle vertical axis proceed two 

 pairs of beaks to both sides, the lower, smaller ones being adorned with ovals. 



On the collar in Fig. 2, Plate xxn, occurs a design of similar style, except 

 that in the under part the two ovals are put side by side between the two out 

 stretched heads. The border ( Fig. 2 a), made for the same robe as the collar, is 

 embroidered with a double pattern. In that to the left the wave-line is formed 

 likewise of cock-heads, between which are confronting conventionalized fishes of 

 two different forms alternating with each other. 



In the following embroideries the single parts constituting cock and fish 

 ornaments are more or less torn apart, displaced, and partly distorted ; so that it 

 is hard to define in every case exactly what represents a cock or a fish, or where 

 the beginning or ending of these creatures is. We see, for instance, in Figs. 3, 4, 

 43, and 5, Plate xxn, simple and compound triskeles in various styles and com 

 binations, grouped together with spirals. A comparison with the forms hitherto 

 described undoubtedly proves them to be derived from components of the cock. 

 The high degree of distortion gives so much individual freedom of choice as to 

 interpretation, considering the ambiguity of the significance of the single pieces, 

 that it would be a hopeless task not only for the Western student of these orna 

 ments, but also for the cleverest native connoisseur, to draw any conclusion as to 

 the details of this ornament. At this point a geometrical stage opens up, where 

 realistic explanation is hopeless, and beyond the pale of which no one can go. 

 That there is, however, an undeniably close continuity between these various 

 degrees of evolution is evidently shown by Figs. 6, 6 a, 6b, all patterns belonging 

 to the same garment. While, as regards Fig. 6, we can but feel like declaring 

 our non possumns, still we are able to decipher the two cocks with their downward- 

 bent beaks and oval bodies in Fig. 6 a, and even the two conventionalized fishes 

 placed together in a figure the geometrical character of which seems to be strongly 

 emphasized, at first sight, in the upper part of this ornament. In Fig. 6b it might 

 be possible to distinguish the cock-beaks, through the circles placed in front of 

 them, on the spiral to the left, as well as the cock filling the concavity of the fol 

 lowing wave-line ; but in this case it is next to impossible to state with certainty 

 which part is to be looked upon as head or tail, granting that these two possibili 

 ties are admissible. 



The preceding remarks apply also to Figs. 1-3, Plate xxm. Only the two 

 combatant cocks over the last pair of spirals in Fig. 2, and the two conventional 

 ized fishes turned away from each other in Fig. 3, may be recognized as such 

 with any degree of certainty. 



In the following designs a definite group of ornamentations is exhibited. 

 There is a double principle active in them, that of displacement and that of 

 combination. 



Fig. 4, Plate xxm, shows a silk collar. The design consists of two ele- 



