550 W. THALBITZER. 
It is only in one drum-song, no. 196, that we have a satisfactory 
basis for a test, namely two phonograms of two singers, and the 
same text. Here a relative similarity is apparent; the two singers 
have had the same aim, but there are many individual variations, 
probably due to the decided rubato execution.‘ 
Unfortunately we lack sufficient material to be able to prove in 
this way that in the case of drum-songs — at all events in the case of 
some of them — the rule holds good that common fixed melodies do not 
exist, only individual melodies. Each singer, then, chooses his own. 
"separate melody-form, either by _composing it himself or by imitating — 
the singers of old, by taking up one of the inherited melodies, or in 
“some other way. However, by another method, the internal criteria 
of certain texts (for instance, in nos. 182—183 and 185—186), we 
find it probable that the drum-contest singer works out his text be- 
fore he is to use his song in the suit, and no doubt at the same 
time has the melody as well in mind. 
Improy isation is doubtless extremely rare i in this art. The drum- 
singer 1S referred to as the owner of the song he has made (pisia, pia), 
that is to say he has a monopoly of, the song with its text and, 
melody аз long as _ he lives. Only after his death is it considered 
suitable for others to use the particular some and melody which 
were his. "That's Akernilik’s or Ak"ko's pisia’, one or another would 
cry, as soon as they heard these songs from my phonograph; the 
_merest fragment of the melody was sufficient for recognition. (Cf. 
p. 167,) 
4. 
Aawtaaritaa’s drum-songs from the Lape Farewell. district have 
faithfully retained the same tonal and rhythmic movements ts which 
we know from Ammassalik. The old tones heard along the east coast 
in a pagan past, filled in 1914 the air of the southernmost district 
of Greenland, which is so closely related to the culture of Ammas- 
salik in spite of being separated by six degrees of latitude and very 
difficult lines of communication. The distance is illustrated by the 
dialect belts, described by Kuannia, between Ammassalik and Cape 
Farewell (see p. 152). These bear witness to a cultural cleft between 
the inhabitants of each place, of many generations’, nay of many 
centuries’ standing. All the drum-songs of the southern corner, and 
1 Nos. 22 and 26 do not count in the question of variants, as the respective 
melodies are only fragments, the beginning and ending of each text respectively; 
melodies 13 and 14, belonging to text no. 22 and sung by one person, should 
have been grouped under one number instead of two. The four DD variants 
under no. 25 are too fragmentary to teach us anything. 
