units on the
basis of "Groups of Hundred", "Groups of Thousand", etc., were created
and the original tribes were dissolved into military regiments. In the
course of time, and especially at the time of the dissolution of a
federation, these military units had gained social coherence and
appeared to be tribes again; we are probably correct in assuming that
all "tribes" which we find from this time on were already "secondary"
tribes of this type. A secondary tribe often took its name from its
leader, but it could also revive an earlier "primary tribe" name.

The Toba represented a good example for this "cone" structure of
pastoral society. Also the Hsiung-nu of this time seem to have had a
similar structure. Incidentally, we will from now on call the Hsiung-nu
"Huns" because Chinese sources begin to call them "Hu", a term which
also had a more general meaning (all non-Chinese in the north and west
of China) as well as a more special meaning (non-Chinese in Central Asia
and India).

The Tibetans fell apart into two sub-groups, the Ch'iang and the Ti.
Both names appeared repeatedly as political conceptions, but the
Tibetans, like all other state-forming groups of peoples, sheltered in
their realms countless alien elements. In the course of the third and
second centuries B.C. the group of the Ti, mainly living in the
territory of the present Szechwan, had mixed extensively with remains of
the Yueeh-chih; the others, the Ch'iang, were northern Tibetans or
so-called Tanguts; that is to say, they contained Turkish and Mongol
elements. In A.D. 296 there began a great rising of the Ti, whose leader
Ch'i Wan-nien took on the title emperor. The Ch'iang rose with them, but
it was not until later, from 312, that they pursued an independent
policy. The Ti State, however, though it had a second emperor, very soon
lost importance, so that we shall be occupied solely with the Ch'iang.

As the tribal structure of Tibetan groups was always weak and as
leadership developed among them only in times of war, their states
always show a military rather than a tribal structure, and the
continuation of these states depended strongly upon the personal
qualities of their leaders. Incidentally, Tibetans fundamentally were
sheep-breeders and not horse-breeders and, therefore, they always
showed inclination to incorporate infantry into their armies. Thus,
Tibetan states differed strongly from the aristocratically organized
"Turkish" states as well as