I do not ask whether, as a matter of fact, this deduction is right or wrong. I only state, as a matter of philological history, that it is made, adding that the hypothesis which makes it is illegitimate. It rests on the assumption that it is easier to bring a population from India to Russia than to take one from Russia to India. In the case of the more extreme language of which it takes cognisance this postulate becomes still more inadmissible. It assumes, in the matter of the Keltic (for instance), that it is easier to bring the people of Galway from the Punjab, than the tribes of the Punjab from Eastern Europe. In short, it seems to be a generally received rule amongst investigators, that so long as we bring our migration from east to west we may let a very little evidence go a very long way; whereas, so soon as we reverse the process, and suppose a line from west to east, the converse becomes requisite, and a great deal of evidence is to go but a little way. The effect of this has been to create innumerable Asiatic hypotheses and few or no European ones. Russia may have been peopled from Persia, or Lithuania from Hindostan, or Greece from Asia, or any place west of a given meridian from any place east of it—but the converse, never. No one asks for proofs in the former case; or if he do, he is satisfied with a very scanty modicum: whereas, in the latter, the best authenticated statements undergo stringent scrutiny. Inferences fare worse. They are hardly allowed at all. It is all "theory and hypothesis" if we resort to them in cases from west to east; but it is no "theory" and no "hypothesis" when we follow the sun and move westwards.
Let the two lines be put on a level, and let ethnographical philology cease to be so one-sided as it is. Let the possibility of a Western origin of the Sanskrit language take its natural place as the member of an alternative hitherto ignored. I do not say what will follow in the way of historical detail. I only say (in the present paper at least) that the logic of an important class of philological questions will be improved. As it stands at present, it is little more than a remarkable phenomenon in the pathology of the philological mind, a symptom of the morbid condition of the scientific imagination of learned men.
Turning westwards we now take up the Slovenians of Carinthia and Styria on their western frontier, not forgetting the southermost of the Czekhs of Bohemia. How far did the Slavonic area extend in the direction of Switzerland, Gaul, and Italy?
In the Tyrol we have such geographical names as Scharn-itz, Gshnitz-thal, and Vintsh-gau; in the Vorarlberg, Ked-nitz and Windisch-matrei. Even where the names are less definitely Slavonic, the compound sibilant tsh, so predominant in Slavonic, so exceptional in German, is of frequent occurrence. This, perhaps, is little, yet is more than can be found in any country known to have been other than Slavonic.
Again—a Slavonic population in the Vorarlberg and Southern Bavaria best accounts for the name Vind-elicia.
If the Slavonians are aboriginal, and if the Czekhs are the same, the decisive evidence that, within the historical period, they have both receded is in favor of their respective areas having originally been greater than they are at present. Such being the case, we may bring them both further south and further west. How far? This is a question of minute detail, not to be answered off-hand. The rule of parsimony, however, by which we are forbidden to multiply stocks unnecessarily, carries them to the frontier of the Gauls in one direction, and the Italians on the other.
If so, there may have been Slavonians on the frontier of Liguria. More than this the Rhæti may have been Slavonic also. But many make the Etruscans Rhætian. Is it possible however, that even the Etruscans were Slavonic?
I know of numerous opinions against their being so. I know of no facts.