142 



supposed lo lie on both sides of the unbroken line. The line 



L 1 L ■: L 

 for such a word as [aju.raîi'eqivq] would look like this : 



which corresponds to the first type. I am not so sure, however, 

 that a slight rise will not often take place at the end here (and 

 in similar cases), already in tlie vowel a, and rapidly fading 

 away in the unvoiced q without having attained to the height 

 of pitch of the preceding syllable ; if this is the case, we have 

 here too one of the last two types. It is certain, at any rate, 

 that it is not necessary for a period to be interrogative to 

 have the rise of pilch at the end, just as, on the other hand, 

 it is not absolutely necessary for a period to have such a 

 rise at the end in order to be interrogative. Yet as a rule, 

 questions have the rise in pitch. Typical examples of the tone- 

 movements in Greenlandic are, in my estimation, such as those 



1111 11 11 



in \simp•гse\'^ [susa-Tpotfci]'-. There is a constant tendency to 



let the voice change in pitch from syllable to syllable with 



ascending intervals. The whole melody of the voice has, as it 



were, a regular type, which agrees with this fundamental rule. 



The compound intonation is not at variance with this tendency; 



if there should arise any disagreement between the two through 



the collocation of the parts of the word or the sentence, a 



kind of tone-assimilation takes place within the word or the 



sentence, whereby the pitch of the syllables is adapted to the 



type (as when ^ ^ "i is changed to s i. 



IS L S 1 L 



Such a formation as [sumck-umuhyn-cty^ 'makes an ora- 



11 L 1 L 1 



torical impression by the side of the natural [sumck-uniahy ir ct\. 



* are you sleeping? '' I did not catch anything. ' wherever I may 

 happen 10 be lin the future). 



