360                          ALBERUNTS INDIA.

eras it is the custom to reckon only with complete, not
with incomplete or current years."

Then the author goes on to give rules for the comparison
of the Saka era with the Greek, Persian, and Muslim eras.

A later author, 'Abii-Sa'id 'Abd-alhayy Ibn Aldahl.iak
Ibn Mahmud Gardezi (Gardez, a town east of Ghazna),
has reproduced the information of Alberuni on the Saka
era in Persian. Not having the original (MS. Ouseley 240,
Bodleian Library, Oxford) at my disposal, I give a trans¬
lation made years ago :—

"The Hindu era is calledji^^ because J l^ (kdla) means
time, and <jj>,^ (Saka,) is the name of a king whose death
was made an era ; he did the Hindus a great deal of harm,
so they made the date of his death a festival" (Oxford
manuscript, p. 352).

The place Ketritr is also mentioned in the (Jhaehndmet.
Vide Elliot, "History of India," i. 139, 143, 207.

P. 7. Al-arkctml.—Gf. note to i. 312. The book does
not seem to exist in the collections of Arabic manuscripts
in Europe.

P. 8.—The pronunciation of the names Kanir, Bardari,
Marigala, and Nirahara (Nlra-griha ?) is more or less con¬
jectural.

Alberuni identifies Mdrigetlet with Taks]ta,silct (vol. ii.
302), i.e. the Taxila of the ancients. The name Mdrigalct
seems to be preserved in that of a ra,ngc of hills lying only
two miles to the south of Shahdhesi (Cunningham, " Ancient
Geography of India," p. 111). The place is also mentioned
in the Tdbaketti-Ndsiri. Vide Elliot, "History of India,"
ii. 271, 273.

P. 9.—Durletbhet, a native of Multan, is only twice men¬
tioned. Here the author quotes from him a method for
the computation of the Saka era, and p. 54 a method for
the computation of cthctrgctnct. According to him, the
Indian year commenced with the month Margasirsha, but
the astronomers of Multan commenced it with Caitra
(p. 10).

P. 10. Betrhalctkin.—The name occurs only in this one