64 [96] TRANS. ST. LOUIS ACAD. SCIENCE. 



quently lived on friendly terms with the Kasi'hta. Kolusa Hatchi 

 is supposed by Dr. Brinton to stand for Tuskalusa Hatchi, the 

 Cha'hta name for Black Warrior river, a laige affluent of the 

 Tombii^bee river. The headwaters of this river really lie in the 

 track of the assumed migration ; but the text distinctly states^ 

 that the stream was called so " because it was rocky there and 

 smoked." I know of no etymology which could explain the term 

 in giving it this signification. 



The ancient and renowned town of Kusa (q.v.) was the next 

 station which the Kasi'hta reached on their eastern journey. 

 Kusa has always been considered as an ancient capital and cen- 

 tre of the Maskoki people ; it appears as such in the documents 

 extending from the i6th to the end of the iSth century. Accord- 

 ing to our legend, the settlement of Kusa is older than the Kasi'- 

 hta immigration ; which means to say, that Creek Indians lived 

 in the Creek country long before the Kasi,|ta arrived there. The 

 cause why Kusa is mentioned here at all, is that they obtained 

 there the war-fetish of the istipdpa ; bones, having previously 

 obliged the Kusa Indians by killing that dangerous "lion" in the 

 ingenious manner described. 



The local names now become more frequent. After leaving the 

 friendly Kusa, the Kasi'hta reached Nofapi or " Beech-tree" riv- 

 er, now called Kalasi Hatchi. The former name belongs to the 

 Hitchiti dialect, and in Creek would sound Nifapi (all the vowels 

 pronounced short) ; Kalasi hatchi, of uncertain etymology, was 

 the name by which the stream was known in Tchikilli's time. 

 This shows that the migrating tribe had then arrived within 

 the limits of the country where some dialect of the Southeastern, 

 Hitchiti or Apalachian branch was spoken at that time, though 

 at a later period the use of Creek predominated there. A pas- 

 sage occurring in I. Gerar W. de Brahm* shows in which part 

 of Alabama we have to look for this stream : 



"Nophabee, Tukasatchee and Tallesee are towns between the first and 

 second cataract [of Locushatchee, now Talepusee river]." Tukasatchee 

 being misspelt for Tukabatchi in that document, which was written early 

 in the i8th century, and is full of inaccuracies of this kind, we see at once 

 that the "Beech creek" or Nofapi must be in the vicinity of Tukabatchi and 



* "Hist, of the Prov. of Georgia," Wormsloe, 1S49 (fol.) p. 55. 



