COMPARATIVE PHONOLOGY OF FOUR SIOUAX LANGUAGES. 923 



which Dr. Kiggs said, "These may possibly be forms of contraction, 

 but we have not now the means of showing the fact." 



Almost any sound can be prolonged in (pegiha and ^oiwere, if it be 

 alone, or stand at the end of a syllable. Propagation occurs in many 

 interjections, and is often used for emphasis. E. g. ha n ega n tce, the 

 morning ; ha n ega n tc8qtci, very early in the morning; ha n +ega n tceqtci, 

 ve — ry early in the morning! Weahide,/ar; we+ahide,/a — r! 



Sound shifting. — As far back as 1872, the writer began to notice many 

 examples of sound-shifting, while comparing the Dakota with the Ponka. 

 After a study of the jr>iwere and Hotcangara, begun at the Omaha and 

 Winnebago reservations, Nebraska, in 1878, he began the formulation 

 of the principles discovered. One of the most important ones is that 

 of triliteral syllables. A triliteral monosyllable in .xoiwere (and often 

 the corresponding ones in Dakota and (pegiha) is changed into a quad- 

 rilateral dissyllable in Hotcaiigara, when the first letter of the mono- 

 syllable is a mute, a palatal spirant, or a spirant sibilant; and the sec- 

 ond consonant is a labial or dental mute, or a dental spirant. The first 

 consonant of the Hotcangara dissyllable is always a surd, the second is, 

 as in the corresponding j^oiwere word, a labial or dental mute, or else a 

 dental spirant; and each consonant (in Hotcangara) must be followed 

 by the same vowel sound. In no case, as far as examples have been 

 gained, can any mule stand next to one of the same order; e. g., a labial 

 cannot precede a labial. 



It is probable that the Dakota biliteral monosyllables "da" and "du" 

 were originally triliteral syllables, an initial "c" having been dropped. 

 That is, Nos. 32, 33, 34, 36, and 37 of the Dakota column in Table HE 

 were originally cda, cda-tka n , cda, cdu, and cduxo, respectively. This 

 seems very probable when we find clo (Teton for cdo), in No. 35, equiv- 

 alent to the (pegiha onu, and cda (No. 31) equivalent to the (fegiha ona. 

 In No. 58 there seems to have been a transposition of syllables, final -ha 

 being equivalent to (pegiha initial a, and initial du to (fegiha final oni n . 

 So, Dakota mdu, equivalent to (pegiha b^i, and mdu-ha to (pegiha a-b$i n ; 

 Dakota yu to (pegiha <£i, and yu-ha to $egiha a-<£i n . 



In comparing the Ponka with itself, or with the Dakota, we find in- 

 stances of permutations of sound, as follows: c and q, x and z, <£ and n. 

 The words in which these occur are not always synonyms, but when we- 

 find a word in which " c," for example, is used, we may infer that the 

 language contains another word, differing from it only in the substitu- 

 tion of "q" for "c." Or, one language uses "c" where the other em- 

 ploys "q." 



The meanings of the words and syllables in the following table will 

 be found at the end of this article. 



