cc
to the -6* and -<22 in Gilgit, Bamiyan, and North Indian documents,
where the point is turned to the lower left side . Both shapes
( 202 are to be found in the Khotanese ms. (a3 ^22^^2*2222222^), while the
1 20i §or6uq ms. has preserved only the Indian form. All mss. contain-
ing Central Asian -f and -222 have been termed by myself "Early
Turkestan Brahmi"^. Like most such terms, this is somewhat
artificial. The Sanskrit fragment from §or^uq, which does not
contain Central Asian -<? and -222, cannot have been written much
earlier than the Khotanese one. But it is classified in accordance
with the definition as belonging to the "Turkestan Gupta type".
Nevertheless, Central Asian and -222 are key features of the
northern route as already stated by A.F.R. HOERNLE (first
mentioned 1893: 4). Among the southern Turkestan types of
Brahmi script they are found only in the earlier mss. Later on,
the Indian forms of and -222 predominate again (PI. 204 ^).
Khotanese material written in "Early Turkestan Brahmi" is com-
paratively rare. It is followed chronologically by three more or
less different types. According to my terms for the Brahmi of the
northern Silk Road, I name them: "Early South Turkestan
Brahmi", "South Turkestan Brahmi", the main type, and "Late
South Turkestan Brahmi". Some characteristics of these Brahmi
types will be discussed later.
Unlike the northern rim of the Tarim basin, where the Brahmi
tradition existed uninterrupted from the Kusana period (SAN-
DER 1968: 2—6, 43—47), on the southern route the Brahmi
script was most probably not used before the fifth century A.D.
It only slowly drove back the Kharosthi script (BERNHARD
1970: 55—62). The events leading to this change are obscure
(EMMERICK 1979a: 1-5; GROPP 1974: 31). It was assumed
by F. BERNHARD that the HInayana tradition of the Dharma-
guptaka school, using Northwest Prakrit or Gandhari as their lan-
guage and Kharosthi as their script, may have sent the first Bud-
dhist missionaries to Central Asia. It may be that the increasing
popularity of the school of the Sarvastivadins/Mulasarvastivadins
from the Kusana era onwards, using Sanskrit as their language and
Brahmi as their script, had driven back the Dharmaguptakas. At
the same time the Mahayana movement became popular and
spread to Khotan. Approximately from the fifth century onwards,
Sanskrit texts were translated into Khotanese, which were the
114
to the -6* and -<22 in Gilgit, Bamiyan, and North Indian documents,
where the point is turned to the lower left side . Both shapes
( 202 are to be found in the Khotanese ms. (a3 ^22^^2*2222222^), while the
1 20i §or6uq ms. has preserved only the Indian form. All mss. contain-
ing Central Asian -f and -222 have been termed by myself "Early
Turkestan Brahmi"^. Like most such terms, this is somewhat
artificial. The Sanskrit fragment from §or^uq, which does not
contain Central Asian -<? and -222, cannot have been written much
earlier than the Khotanese one. But it is classified in accordance
with the definition as belonging to the "Turkestan Gupta type".
Nevertheless, Central Asian and -222 are key features of the
northern route as already stated by A.F.R. HOERNLE (first
mentioned 1893: 4). Among the southern Turkestan types of
Brahmi script they are found only in the earlier mss. Later on,
the Indian forms of and -222 predominate again (PI. 204 ^).
Khotanese material written in "Early Turkestan Brahmi" is com-
paratively rare. It is followed chronologically by three more or
less different types. According to my terms for the Brahmi of the
northern Silk Road, I name them: "Early South Turkestan
Brahmi", "South Turkestan Brahmi", the main type, and "Late
South Turkestan Brahmi". Some characteristics of these Brahmi
types will be discussed later.
Unlike the northern rim of the Tarim basin, where the Brahmi
tradition existed uninterrupted from the Kusana period (SAN-
DER 1968: 2—6, 43—47), on the southern route the Brahmi
script was most probably not used before the fifth century A.D.
It only slowly drove back the Kharosthi script (BERNHARD
1970: 55—62). The events leading to this change are obscure
(EMMERICK 1979a: 1-5; GROPP 1974: 31). It was assumed
by F. BERNHARD that the HInayana tradition of the Dharma-
guptaka school, using Northwest Prakrit or Gandhari as their lan-
guage and Kharosthi as their script, may have sent the first Bud-
dhist missionaries to Central Asia. It may be that the increasing
popularity of the school of the Sarvastivadins/Mulasarvastivadins
from the Kusana era onwards, using Sanskrit as their language and
Brahmi as their script, had driven back the Dharmaguptakas. At
the same time the Mahayana movement became popular and
spread to Khotan. Approximately from the fifth century onwards,
Sanskrit texts were translated into Khotanese, which were the
114