No. 5.] CAMPBELL — HITTITES IN AMERICA. 277 



tite confederacy embraced within it all these diflferent features. 

 From what source then has Dr. Hyde Clarke obtained materials 

 for a comparison of the Khita with the Peruvian ? That source 

 is the Accad, or ancient language of the primitive inhabitants of 

 Chaldea, vocabularies of which are preserved in Assyrian tablets, 

 tosrether with bilin";ual records and treatises. The Accadians 

 were undoubtedly a Turanian people, the predecessors of the 

 Semitic occupants of the Tigro-Euphrates basin, and their lan- 

 guage bears a well defined Turanian stamp. Assyriologists 

 generally refer it to the Ugrian family as kindred to the Lapp, 

 Finn, Magyar, etc. Still the Accad differs from the other mem- 

 bers of this family in its constructions. Like them, it employs 

 postpositions and postpositional pronouns, and places the verb 

 after its regimen. But, unlike these languages, it places the 

 nominative before the genitive and the adjective after the noun, 

 as do the Celtic dialects. In the postposition of the genitive, it 

 also differs from the Khita language as indicated by its few 

 remains. But the Khita sar is a thoroughly Accad word, and 

 Ashtar, the god of the Khita, can be no other than Hasisadra 

 -of the Accadian mythology. Instead, therefore, of employing 

 the term Accad, Dr. Hyde Clarke takes the word Khita as 

 more comprehensive, being convinced of the essential unity of 

 the Accadian and Hittite populations. 



It is worthy of note, however, that the ancient rulers of Chal- 

 dea termed themselves " Kings of Sumer and Accad," and 

 reference is constantly made to this double constitution of the 

 monarchy, as important apparently as the later distinction be- 

 tween the Medes and Persians, as elements in one nationality. 

 True there survives no Sumerian grammar or dictionary as dis- 

 tinguished from the Accadian, so that full license has been 

 afforded to philologists to denote the language by either name, 

 or to suppose that one of them, Sumer or Accad, was a Semitic 

 dialect and the parent of the later Chaldean. From the double 

 character of Accadian grammar, as partly Turanian and partly 

 Semitic or Celtic, from the presence of a large number of purely 

 Celtic words in the language alongside of others as purely Tura- 

 nian, and from the very name Sumer itself as related to Kymri, 

 with other facts which will come out in the sequel, I am com- 

 pelled to the conclusion that Accad, as we possess it, is a 

 compound language, in which Khita or Accad proper exists in 

 union with Celtic or Sumerian, both as regards grammatical and 



