The racial affinities of Mitanni are doubtful. Prince, correspondence, says the language of Mitanni was certainly not Aryan. It has been thoroughly analyzed by Ferdinand Bork, in his _Die Mitanni Sprache_, who compares it with the Georgian or Imeretian branch of the Caucasic linguistic groups. The Mitanni are not to be confused with the Ossetes, who speak a highly archaic, real Aryan language. Mitanni, in structure, is like the polysynthetic North American groups. Feist, 1, p. 14, says the Mitanni were Nordics and inhabited the western mountains of Iran, in Zagros. In 5, p. 406, he places them on the north of the Euphrates during the fifteenth and sixteenth centuries B. C. See also Hall, p. 200, the following note and that to p. 213 : 1–23 of this book. Hall also considers them Nordics. 239 : 16 _seq._ Von Luschan, p. 230, asks: “Can it be mere accident that a few miles north of the actual frontier of modern Kurdish languages there is Boghaz-Köi, the old metropolis of the Hittite Empire, where Hugo Winckler in 1908 found tablets with two political treaties of King Subiluliuma with Mattiuaza, son of Tušrata, king of Mitanni, and in both of these treaties Aryan divinities, Mithra, Varuna, Indra and Nasatya are invoked, together with Hittite divinities, as witnesses and protectors? And in the same inscriptions, which date from about 1380 B. C., the king of Mitanni and his people are called Harri, just as nine centuries later in the Achæmenidian inscriptions Xerxes and Darius call themselves Har-ri-ya, ‘Aryans of Aryan stock.’ So the Kurds,” concludes Von Luschan, “are the descendants of Aryan invaders and have maintained their type and their language for more than 3300 years.” See also the notes to p. 173 : 11. 239 : 29. See pp. 128 and 137 of this book. 240 : 4 _seq._ See the notes to p. 173. 240 : 15 _seq._ See the notes to p. 242 : 5. CHAPTER XIII. ORIGIN OF THE ARYAN LANGUAGES 242 : 5. The following notes on languages were taken mostly from the _History of Language_, by Henry Sweet, and were supplemented by the writings of W. D. Whitney, and an article on “Indo-European Languages,” by Peter Giles. All languages may be roughly divided into two great groups, _isolating_ and _agglutinative_. The isolating languages are constructed on the principle of single, distinct words for each idea, and do not employ forms which add or drop syllables, or letters, in order to obtain variety of expression, tense, mode, person, number, etc. However, the element of intonation frequently plays a large part in multiplying the number of possible forms, and therefore of ideas, in isolating languages, by imparting to otherwise identical words different meanings through pitch, rising or falling inflection or accent.