was still absent.  The main obligations of these lords were to send tributes of grain, to participate with their soldiers in the wars, to send tortoise shells to the capital to be used there for oracles, and to send occasionally cattle and horses.  There were some thirty such dependent states.  Although we do not know much about the general population, we know that the rulers had a patrilinear system of inheritance.  After the death of the ruler his brothers followed him on the throne, the older brothers first.  After the death of all brothers, the sons of older or younger brothers became rulers.  No preference was shown to the son of the oldest brother, and no preference between sons of main or of secondary wives is recognizable.  Thus, the Shang patrilinear system was much less extreme than the later system.  Moreover, the deceased wives of the rulers played a great role in the cult, another element which later disappeared.  From these facts and from the general structure of Shang religion it has been concluded that there was a strong matrilinear strain in Shang culture.  Although this cannot be proved, it seems quite plausible because we know of matrilinear societies in the South of China at later times.

About the middle of the Shang period there occurred interesting changes, probably under the influence of nomad peoples from the north-west.

In religion there appears some evidence of star-worship.  The deities seem to have been conceived as a kind of celestial court of Shang Ti, as his “officials”.  In the field of material culture, horse-breeding becomes more and more evident.  Some authors believe that the art of riding was already known in late Shang times, although it was certainly not yet so highly developed that cavalry units could be used in war.  With horse-breeding the two-wheeled light war chariot makes its appearance.  The wheel was already known in earlier times in the form of the potter’s wheel.  Recent excavations have brought to light burials in which up to eighteen chariots with two or four horses were found together with the owners of the chariots.  The cart is not a Chinese invention but came from the north, possibly from Turkish peoples.  It has been contended that it was connected with the war chariot of the Near East:  shortly before the Shang period there had been vast upheavals in western Asia, mainly in connection with the expansion of peoples who spoke Indo-European languages (Hittites, etc.) and who became successful through the use of quick, light, two-wheeled war-chariots.  It is possible, but cannot be proved, that the war-chariot spread through Central Asia in connection with the spread of such Indo-European-speaking groups or by the intermediary of Turkish tribes.  We have some reasons to believe that the first Indo-European-speaking groups arrived in the Far East in the middle of the second millennium B.C.  Some authors even connect the Hsia with these groups.  In any case, the maximal distribution of these people seems to have