with an orthodox equal. However, in 525, Ts'u and Wu "battle" together, the commentator explaining that Ts'u is now "promoted" to battle rank, though the strict rule is that two barbarians, or China and one barbarian, "beat" rather than "battle." In 591 Confucius had already announced the "end" of the King of Ts'u, not as such, but as federal viscount. Under ordinary circumstances "death" would have been good enough: it is only in speaking of his own ruler's death that the honorific word "collapse" is used. All these fine distinctions, and many others like them, hold good for modern Chinese. These (apparently to us) childish gradations in mere wording run throughout Confucius' book; but we must remember that his necessarily timid object was to "talk at" the wicked, and to "hint" at retribution. Even a German recorder of events would shrink from applying the word _haben_ to the royal act of a Hottentot King, for whom _hat_ is more than good enough, without the _allergnädigst._ And we all remember Bismarck's story of the way mouth-washes and finger-bowls were treated at Frankfurt by those above and below the grade of serene highness. _Toutes les vices et toutes les moeurs sont respectables._ In 531 the barbarian King of Ts'u is honoured by being "named" for enticing and murdering a "ruler of the central kingdoms." The pedants are much exercised over this, but as the federal prince in question was a parricide, he had a _lupinum caput,_ and so even a savage could without outraging orthodox feelings wreak the law on him. On the other hand, in 526, when Ts'u enticed and killed a mere barbarian prince, the honour of "naming" was withheld. This delicate question will be further elucidated in the chapter on "Names." It will be observed that none of the testimony brought forward here to show that Ts'u was, in some undefined way, a non-Chinese state is either clear or conclusive: its cumulative effect, however, certainly leaves a very distinct impression that 'there was a profound difference of some sort both in race and in manners, though we are as yet quite unable to say whether the bulk of the Ts'u population was Annamese, Shan, or Siamese; Lolo or Nosu; Miao-tsz, Tibetan, or what. There is really no use in attempting to advance one step beyond the point to which we are carried by specific evidence, either in this or in other matters. It has been said that no great discovery was ever made without imagination, which may be true; but evidence and imagination must be kept rigidly separate. What we may reasonably hope is that, by gradually ascertaining and sifting definite facts and data touching ancient Chinese history, we shall at least avoid coming to wrong positive conclusions, even if the right negative ones are pretty clearly indicated. It is better to leave unexplained matters in suspense than to base conclusions upon speculative substructures which will not carry the weight set upon them. CHAPTER XXVIII BARBARIANS