The bolder doctrine is the assumption of what has been called The Thraco-Pelasgic stock. This maintains that the extinct populations and languages of Thrace, Mœsia, and Pannonia were intermediate to those of the two peninsulas, and that, by a sort of divarication, the western extension of their southern members peopled Italy, and the eastern, Greece. This view has the advantage of being difficult to refute--since it is the current belief that the original languages of the three countries in question are extinct, and that, as nothing is known about them, it is as easy to say that they were the mother tongues of the Greek and Latin as aught else. The assumed displacements, however, are enormous; besides which, the ancient Thracians must have been more Greek than were the ancient Italians; which is unlikely. But the great difficulty in fixing a locality for this Thraco-Pelasgic, or Helleno-Latin language (call it what we will) lies in a reason which the reader of the first chapter of this book may, perhaps, anticipate. It lies in the existence of the Albanian language; a fact, which I said, on the onset, was one of such importance as to require being treated as a special and separate preliminary to the ethnology of Greece and Italy, as well as on its own merits. Whence came this remarkable tongue, and whence the populations who speak it? For a long time both were considered recent introductions,--introductions from Caucasus, perhaps, or from some other locality equally plausible. But this origin is no longer admitted by any competent investigator; and the modern Skipetar, or Albanians, are now looked upon as the descendants of the ancient Illyrians, and of such Epirots as were not truly Greek. So that the Thraco-Pelasgic hypothesis is materially weakened by the inconvenient locality, and the impracticable antiquity of this nation. So awkwardly does it lie, that it fills up full two-thirds of the area required for the hypothetical tongue in question. Hence the line of such transitional populations as, by connecting Greece and Italy, account for the ethnological affinities of their respective occupants, must not be a straight one. On the contrary, it must trend _round_ the Albanian country, _viâ_ Macedon, Thrace, Servia, Croatia, and Carniola. The assumption of a stream of population from Asia Minor across Turkey, Servia, and the parts to the north of the Adriatic is the Thraco-Pelasgian doctrine modified; since it deduces both tongues from a common source. The assumption of a similar stream across the islands of the Ægean does the same. Yet each is beset with difficulties. If one fact be better supported than another, it is that the Ægean islands and the Asiatic coast were peopled from Greece rather than _vice versâ_.