But the more conspicuous part of the decoration is formed by bands of tall cufic[111] letters and by flowers, more or less schematised. Apart from a fleur-de-lis, which occurs chiefly in medallions, the most important flower is the Oriental lotus. This flower as it appears relieved on a blue ground in the later mosque lamps is identical in drawing with the lotus that we see so frequently in Indian and Chinese art. It is often combined with what at first sight appears to be another flower, treated _en rosette_, with an involucre of six oval and six triangular petals, and an indication of a seed-vessel in the centre; but this again may perhaps be only the same lotus-flower seen full-face. In some cases, as on certain mosque lamps, these flowers, broadly treated, form the sole decoration; but more often the floral design passes into the formal schematised patterns so characteristic of Arab art at this time. [Illustration: _PLATE XXIII_ SARACENIC ENAMELLED GLASS THIRTEENTH CENTURY] The medallions that interrupt the broad bands are an essential part of the decoration; they are filled sometimes with inscriptions, generally in this case in the _nashki_ or running script, or more often with certain badges, which are of much interest in connection with the heraldry, if it can be so called, of the day. These badges are derived from the most divergent sources: there is one simple design that resembles the cartouche of an old Egyptian king—it has even been read as ‘Lord of the Upper and Lower Country’ (a good example may be found on a bottle at South Kensington). Another badge takes the form of a strange bird with long tail-feathers, undoubtedly derived from the imperial phœnix of China; any hesitation as to the origin of this design is removed on observing in the field certain little curly clouds, an essentially Chinese motive. A sword, a pair of polo-sticks, or still more often a cup, charged upon a fesse or band which divides the medallion, are badges of more local origin. The same may probably be said of the eagle variously displayed, which in one example, in the British Museum, occurs exceptionally upon an ovoid shield. In some cases the Memlook sultans and emirs adopted ‘canting badges’ based upon their Turki names; as, for example, the well-known duck of the Sultan Kelaoun. The identification, however, of the owner, or the date of a vase or lamp from these badges alone, is, in the absence of an inscription, a somewhat hazardous proceeding. It is a curious fact that we have only two instances of a signature of an artist in all this series of enamelled glass. On a lamp from the Mannheim collection, now, I think, belonging to Mr. Pierpont Morgan, an inscription in running characters on the foot has been read: ‘Work of the poor slave Ali, son of Mohammed Ar Ramaki (?), God protect him’ (Schmoranz, p. 67). It is the same Ali, apparently, who signs his name on another lamp described by Artin Pasha.