s no
greater marvel than that mere children in Paris speak French. But, however
convenient, as an inter-racial and commercial language, Arabic may be to
those accustomed to it, or naturally conversant with it, it is most
difficult to learn by Western races. With ten years' study one may become
a good scholar, and proficiency may follow for the persistent few. This
will explain why inscriptions, texts, and verses on rugs and carpets are
meaningless, except to the most erudite; and except, also, to those who
see in them only another phase of Persian ornament, strange, mysterious,
arabesque, and beautiful.

Regarding the date, often woven into an example which the artist thought
especially worthy, it would seem that some simple formula might be given
for its ready translation. This may be approximated, although it is not so
easy a matter as might appear, and requires a few words on the subject of
Arabian numerical notation. Their general system is similar to ours, and,
corresponding to our miscalled "Arabic figures" of:

0, 1, 2, 3, 4, 5, 6, 7, 8, 9,

their digits are represented by

[Arabic]

Both are read from left to right. These Arabic digits, however, are not
always easily to be deciphered on a rug, on account of the spreading of
the wool and consequent irregularity of outline, and also because they
generally appear in modest size. The back of the rug will show the figures
much more sharply than the face, when there is a doubt. When the
Arabic numerals are made clear, it remains to reduce this date to the
corresponding one of the Christian era, by means of a complicated table.

[Illustration:

PLATE XII.
BERGAMA
_Thirty to forty years old_
FROM THE COLLECTION OF THE AUTHOR
Size: 3.1 x 3.5]

All Mohammedan dating (with exceptions not to be considered here, however
interesting historically) is from the Hegira. The reckoning is not from
the time of Mohammed's "flight" from Medina (September, 622), but from a
day about two months earlier; namely, the first day of that Arabian year.
This beginning of the epoch, according to the best modern authorities,
probably corresponds to July 16, 622.

Mohammedan chronology, however, is often expressed in other ways than by
clear figures, and such florid records are most difficult to interpret.
Again, in old manuscripts, on coins and on a few rare antique carpets, the
date is written out in full, in so many words; as, for instance,
"two-hundred-an