It occurs in the question before us; leading to the inference that if both the Keltic of the Cisalpine Gauls, and the Etruscan of Circumpadane Etrurians were less unequivocally Indo-European than the Slavonic of the Norici and the Umbrian of the Umbri, the original occupants of the intervening area must have been neither Gauls nor Etrurians, but one of four things-- 1. Members of the class to which the Umbrians belonged-- 2. Members of the class to which the Norici belonged-- 3. Partly Norici and partly Umbrians-- 4. Transitional populations sufficiently different from each to constitute a third class, but sufficiently allied to each to be more Norican or more Umbrian than aught else. Such is the way in which here, as elsewhere, we must attempt the reconstruction of what may be called _areas of original connection_. In the present case, then, north-eastern Italy was originally divided between the true Italians, akin to the Umbri, and the extinct or modified Slavonians of Liburnia and the country of the Veneti. Apulia, Calabria, and Sicily, to which we may now attend, I imagine, in the earliest times, to have been occupied _by the ancestors of the Greeks_, a doctrine to which I direct the careful consideration of scholars; since it implies a great change in all our preconceived opinions, and not only makes the Hellenes of Greece as foreign to Hellas, as the Anglo-Saxons were once to England, but deduces them from Italy, and that by means of a _maritime_ migration--a maritime migration which implies not only that they were a population foreign to the Greek soil, but that their descendants were a mixed stock; since no mode of migration is less favourable to the purity of the migrant population than a sea-voyage, where space is limited and females are an incumbrance. Such was, undoubtedly, the origin of the Greeks of Asia Minor and the Ægean Islands. Such, I believe, to have been the origin of the Greeks of Peloponnesus and Northern Hellas. The observations on the relation between the Slavonic and Latin languages have prepared the way to this hypothesis, wherein the necessity of finding a geographical connection between cognate forms of speech recurs.