Page:Folklore1919.djvu/139

Rh the Pisaca tongue, which Grierson has so lately made known to us (1906). Gaster suggests that our dialects in Europe represent the tongue of a 'first wave of a westward movement of an Indian tribe or caste,' making a very short stay in Persia, one somewhat longer in Armenia and settling in Rum or Byzantine Empire about the time of the Comneni (1100-1200). A second wave or clan 'followed in their wake,' and during a sojourn in Armenia sent out offshoots southwards to Syria, Egypt and N. Africa. These two tribes, though of common if remote Hindu ancestry, 'must be kept strictly apart' in our inquiry. Finally nothing more precise can be said than this: an essentially Indian tongue has been enriched (but also corrupted) by loans from the nations among whom the gipsies passed and has been so profoundly modified in grammatic inflexion that in several cases (e.g. English and Serbian) 'barely a skeleton has remained' (G.) It is denied by many that a gipsy from Greece or Rumania can understand one from England or Germany. The loan-words tell us of the route taken and also give hints as to the date of sojourn. Arabic words are entirely wanting, and, in the opinion of some, weaken de Goeje's surmise that the gipsies came via Arabia (v. supra). The words from Persian and Armenian show that they came into contact with these tongues when both had assumed their modern form, i.e. at no very remote date. Nor are the Greek or Slavonic loan-words of an archaic type, gipsies of Germany, Italy, France and England employing to-day quite late Byzantine, indeed modern Greek forms; which enables Gaster to lay down dogmatically that 'they could not have been in Europe much earlier than the date given above' (cent. or ). Largely through Miklosich's studies we can trace step by step the deterioration of a language, only kept in purest form in Greece and Rumania: we can prove that, in Europe at least, it was one and the same, slowly splitting