taken to have a racial significance; or, in other words, it is necessary in every case where either the term Turk or Tatar occurs, to see whether the writer is applying it in its general and sociological acceptation, or in a specific and discriminating ethnic sense. The second conclusion is that the word Moghul, even where it is used in an ethnic sense, is frequently misapplied, and so extended, at certain periods in history, as to comprise many tribes of real Turki race (among others), until large numbers of people who were not of Moghul race came to be called Moghuls. This habit appears to have been prevalent first in the time of Chingiz and his immediate successors, and subsequently during the ascendency of the Chaghatai (or so-called Moghul) dynasty in India. The third conclusion is that the application and significance of all three names—Turk, Tatar, and Moghul—varied at different times and in different countries. It appears to me that a due appreciation of these three points will help to clear up much that has been regarded hitherto as inconsistent, and even contradictory, in the Musulman histories, and has occasioned no little controversy among European writers. That the ethnographic nomenclature of Persian, Turki, and Arabic writers is anomalous, cannot but be granted; but in Asiatic nomenclature what is there that is not anomalous? They had no knowledge of the scientific ethnology that guides the modern European commentator on their works, but merely followed the common speech of the time, and employed the terms that had grown into use among the people around them. In reading their books, therefore, it is futile to look for systematic nomenclature; but if they are read with a due regard to date, locality, and other circumstances, they will seldom be found, I think, to contain actual contradictions; for loose and inaccurate though Asiatics are in some respects—such as in figures, measurements, geographical details, etc.—they are usually remarkably clear on such subjects as blood relationship, family lineage, and racial descent.
But here we must leave the Moghuls, and glance briefly at those original Turks, or Uighurs, who may be regarded as the immediate ancestors of the population of Alti-Shahr (and indeed all Eastern Turkistan) and the main stock of their race. Who the Uighurs were in remote times, and what. was their origin, are speculative questions which need not be investigated here. The best notices of them during early historic times point to their home-land as lying in north-western Mongolia;
