140 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1910. 



the border are suggestive of Cairene and Moorish titles of indifferent 

 quality. The carpet came from a mosque at Aleppo. 



In the same class of debased ornament, derived from finer origi- 

 nals, we may place this carpet. Most of the forms are undecipher- 

 able. A cypress tree is fairly evident ; below it to the left is a queer 

 and somewhat entangled device, which I think we may call a dragon ; 

 still lower down, to the left, is a pair of shapes which faintly resemble 

 those of animals, and to the right of them is an equally faint re- 

 semblance of a long-necked bird — possibly a crane. The carpet is 

 called Persian of the sixteenth or seventeenth century. Another 

 specimen elsewhere is described as being made before the fifteenth 

 century. The scheme of the design is usually found in woven stuffs 

 of the fifteenth century, with details and devices of intelligible 

 beauty. I think that the indications are in the direction of estab- 

 lishing such carpets as the one before us as ambitious attempts by 

 careful weavers far removed from properly drafted designs and 

 relying, therefore, upon elusive memories for their ornamentation 

 which does not call for much admiration. 



The strenuous itinerant instrumentalist seated on the pavement, 

 and diligently twanging the strings of his harp with some rhythm, 

 much discord, and uncertain melody is surely a kind of confrere of 

 these weaA^ers of distorted patterns. 



This slide is from two carpets of Indian manufacture at Malabar, 

 the modest patterns of which — especially that on the right hand — 

 are directly borrowed from inlaid work of the fifteenth century, done 

 by Mohammedans at Broussa, in Asia Minor, and in Venice, etc. A 

 similar style of pattern occurs on carpets made at Tanjore. 



A considerable number of Persian carpets were made in the six- 

 teenth century from designs, of which the leading feature was a 

 covering network or framing. Here this feature is carried out so as 

 to produce a succession and series of interchanging panels, each of 

 which is filled in with plant forms, Tartar clouds, arabesques, or 

 pairs of birds. Schemes of net pattern, but in other variations may 

 be traced in Roman mosaics contemporary with Coptic tapestry 

 weavings, as well as in Byzantine shuttle weavings, thence thej pass 

 into mediaeval European textiles and embroideries, and architectural 

 enrichments, before appearing in carpets. 



Another and simpler example of this scheme of design is shown in 

 this next slide from an Indian or Persian carpet of the sixteenth or 

 seventeenth century. The border is much narrower; the net or frame- 

 work is defined in delicate spiral stems, and at their junctions are 

 variously shaped panels, the netAvork itself does not form such recur- 

 rent panels as in the previous specimen, but is independent of those 

 here shown. 



