_Persian frontier._--Reasons for supposing that tribes other than those of the Turk division ravaged Persia as early as the time of Cyrus, would lie in the incompatibility of any accounts of such invaders with the known facts concerning the Turks. I am not aware, however, that any such incompatibility exists. The names are different. No Sakæ or Massagetæ are known, under such denominations, as Turk tribes. Yet this scarcely constitutes even the shadow of an objection; since native names, and names by which tribes are known to nations other than their own, oftener differ than coincide. _The Caucasian frontier--the frontier of the Don._--Here the reasoning becomes more difficult. An invasion of Persia along the frontier from Bokhara to the Caspian, is an invasion which no existing nation could claim, except the Turk; since it is a rule in ethnological reasoning _to consider every nation as indigenous to the country where it is first found, unless reason be shown to the contrary_. For the parts, however, between the Volga, Caucasus, and the Don (or even Dnieper), there is no such present unity of nation as between the Caspian and Bokhara; and an invasion that burst upon Persia from the north-west, or upon Greece from the north-east, might well be claimed for no less than four great ethnological sections.--1. The Turk. 2. The Slavonic. 3. The Circassian. 4. The Hungarian. I will apply general principles to get at the different probabilities here involved. 1. The nation that invades _both_ Persia and Europe is most probably the nation most intermediate to the two. This is in favour of the Cimmerians having come from the present country of the Nogays, rather than from the Ukraine, or from the Bashkir country, _i.e._, in favour of their being Turk rather than Slavonic or Hungarian. 2. A nation that, within the historical period, has always encroached upon others is more likely to be the invader, in a given instance, than a nation which has not been known so to be in the habit of extending itself. This is in favour of the Cimmerians having been Turks from the Nogay country, rather than Circassians. This is the geographical view. Another method is to take the names of certain invading tribes mentioned in history, and to consider how far they belong to the Turk division, or are to be distributed elsewhere. Here the ethnological method is to begin with the most recent:--