In this very matter of the relation between fresco and mosaic, Kahrieh Jami happens to play a particular part. The mosaics are disposed with such a mastery of composition, there is so wide a range of colour in them, in life and naturalness and sometimes in choice of subject they differ so greatly from better-known mosaics of an earlier period, that some critics have seen in them a fine Italian hand—and one no less fine than that of Giotto, who painted the Arena chapel in Padua about the time Metochites restored this church. Not that any one has gone so far as to ascribe the Byzantine series to Giotto himself, but that the qualities I have mentioned, together with certain similarities of detail, have been ascribed to the revolutionary influence of the Italian series. It is not yet unanimously decided whether the mosaics all belong to the same period. Perhaps we must wait for the evidence of those still hidden in the nave to know whether any of them belong to the time of Mary Ducas. The Russian archæologist Schmitt, who has written the completest monograph on the subject—and who picked enough plaster away in the nave to assure himself that mosaics were still there—assigns the work to the period of Metochites, but surmises it to have been inspired by some Syrian original of the ninth century. Diehl, the eminent French Byzantinist, sees rather in Kahrieh Jami a last revival of Byzantine art, contemporaneous with but not derived from the early Tuscan school of painting. When these savants expressed their opinions neither of them was aware of an odd little fact quite lately established not by a Byzantinist but by a layman who was looking at some photographs of the mosaics. In the photograph of the central bay of the outer narthex he discovered, above a two-handled jar which a servant carries on his shoulder to the marriage at Cana, a date in Arabic numerals—but real Arabic numerals, not the ones we have made out of them. This date is 6811, which in the Byzantine system of chronography is equivalent to 1303. The find was interesting in itself as being the earliest use yet recorded—if I am not mistaken—of Arabic numerals on a public monument. It has a further interest in pointing to the Syrian affiliations of the monastery and in lending colour, however slight, to Schmitt’s theory with regard to the Syrian origin of certain of the mosaics. But it tends more definitely to prove that the mosaics were executed before Giotto’s frescoes in Padua, which could hardly have been begun and much less completed by 1303. [Illustration: Mosaic from Kahrieh Jami: the Marriage at Cana Photograph by Sébah and Joaillier, Constantinople]