REPORT OF THE SECRETARY. 21 



considerable attention to the ancient works of that region, as it is prob- 

 able that there and in western North Carolina is to be fonnd the key that 

 will materially assist in solving the problem of the singular works of 

 Ohio. The results of these exi)lorations are of unusual interest, inde- 

 pendent of their supposed bearing ou the Ohio mounds. 



Mr. J. D. Middleton, who has been a constant assistant since the or- 

 ganization of the division, after completing some investigations begun 

 in southern Illinois, visited western Kentucky for the purpose of inves- 

 tigating the works of that section, but was soon afterwards called to 

 Washington to take part in the ofiHce work. During the month of 

 June he visited and made a thorough survey of the extensive group of 

 works near Charleston, West Va., of which Colonel Norris had made a 

 partial exploration, the latter haviug been prevented from completing it 

 by the sickness which immediately preceded his death. During the 

 same month Mr. Middleton commenced the survey of the Ohio works 

 before alluded to, securing some valuable results in the short time be- 

 fore the close of the year. 



Mr. Gerard Fowke was also engaged for a short time in field work in 

 western Pennsylvania, Ohio, and Kentucky, but was called early in 

 autumn to Washington to assist in office work. 



General field studies. — During October and December Mr. Albert S. 

 Gatschet w^as engaged in gatheriug historic and linguistic data iu 

 Louisiana, Texas, and the portion of Mexico adjoining the Rio Grande, 

 this region coutaining the remnants of a number of tribes of whose lan- 

 guage and linguistic affinity practically nothing is known. After a 

 long search Mr. Gatschet found a small settlement of Biloxi Indians 

 at Indian Creek, 5 or 6 miles west of Lecompte, Rapides Parish, La., 

 where they gain a livelihood as day laborers. Most of them speak 

 English more than their native tongue. In fact about two-thirds of 

 the thirty two survivors speak English only. Tlie vocabulary obtained 

 by him discloses the interesting fact that the Biloxi belong to the Siouan 

 linguistic family. 



Mr. Gatschet heard of about twenty-five of the Tunica tribe still 

 living in their old homes on the Marksville Prairie, Avoyelles Parish, 

 La. An excellent vocabulary was obtained of tlieir language at Le- 

 compte, La., and a careful comparison of this with other Indian lan- 

 guages shows that the Tunica is related to none, but represents a dis- 

 tinct linguistic family. 



He was unable to collect any information in regard to the Karankawa 

 tribe, concerning which little is known except that they lived upon the 

 Texan coast near Lavacca Bay. 



Leaving Laredo County, Texas, he visited Camargo in Tamaulipas, 

 Mexico, finding near San ^Miguel the remnants of the Comecrudo tribe, 

 or, as they are called by the whites, Carrizos. Only the older men and 

 women still remember their language. The full-blood Comecrudos seen 



