It is difficult to say what was the nature of the bond between the different districts which made up the whole body of either Ma-han, Chin-han or Pyön-han. On the one hand we are told that the districts were entirely separate and yet we find Ma-han, as a whole, performing acts that imply some sort of federation at least if not a fixed central government. In fact one Chinese work states that a town named Cha-ji was the capital of all three of the Hans. We must conclude therefore from those and subsequent statements that some sort of central government prevailed, at least in Ma-han. The names of the several kingdoms which composed the three Hans are preserved to us, mutilated, in all probability, by reason of Chinese transliteration, but still useful from a philological and ethnological standpoint. If the reader will glance but casually at the list of these separate districts as given in the appendix, he will see that there was good cause for the division into three Hans. We will point out only the most striking peculiarities here, as this belongs rather to the domain of philology than to that of history. In Ma-han we find seven of the names ending _ro_. We find two or three of the same in Pyön-han but none in Chin-hau. In Ma-han we find fourteen names ending in _ri_ but none in either of the others. In Pyön-han we find ten names beginning with Pyön-jin which is wholly unknown to the other two. In this we also find three with the unique suffix _mi-dong_. In Chin-han we find nine ending in _kan_ and five in _kaye_, which are found in neither of the others. It is hardly necessary to say that these cannot be mere coincidences. In each group we find at least one considerable set of endings entirely lacking in the others. As our own ending _ton_, _ville_, _burgh_, _chester_ and _coln_ have an original significance, so these ending _ro_, _ri_, _mi-dong_, _kan_ and _ka-ya_ have a meaning which should supply us with important clues to the origin of the people of southern Korea. The marked polysyllabism of these names makes it impossible to imagine a Chinese origin for them. It is seldom that a Manchu or Mongol name of a place exceeds two syllables. On the other hand we find in Japan and Polynesia common use of polysyllabic geographical names. A thorough discussion of the subject here would be out of place, but this much must be said, that several of these endings, as _ro_, _pin_ and _kan_, find their almost exact counterpart in the Dravidian languages of southern India, where they mean _village_, _settlement_ and _kingdom_.