128 W. THALBITZER. 
on the first syllable and becomes am'it, or alöq ‘a boot-sole’ in plural 
becomes årut. It is the inner parts of the words, not their endings 
that show irregularities. 
$ 35. — The Eastgreenland grammar has the same classes of in- 
flection as Westgreenland’s, cf. my previous exposition of these in 
my grammatical sketch (Eskimo §§ 18—21). 
The phonetic changes just mentioned naturally put their impress 
upon the forms of inflection in the Ammassalik dialect. The alterna- 
tion of п and { as plural-marks, of т and р as relative-marks, also 
known from West Greenland, in the Ammassalik dialect tend to give 
predominance to the nasal sounds. I have been unable to ascertain 
any regularity in this alternation that only seems to be of an occas- 
ional or individual nature.’ The same applies to the consonantal final 
sounds in absolutive q 7 г which alternate in the same words without 
changing the meaning in the least, e. 5. fupeq = tupén = tupér ‘а tent’; 
likewise is k displaced by 1 and ¢ by п in the words ending in the 
sounds without the slightest difference in meaning. 
The fluctuation seems to depend on the different tendencies of 
the individuals in their pronunciation, and to some extent also to be 
caused by the nature of the language, that is, the influence from the 
quality of the following initial sound. The same applies to the loss 
of final п or t Gti for if) and of the relative-mark m or р. 
The loss of т огр is very common in front of words that begin 
with a consonant, so that e. g. instead of the regular to”q:ip qa:ne 
‘on top of the tent’ ({upéq, in relative to”’q: ip) we get loge qa:ne. I 
found that the relative-ending is used only very rarely in the Am 
massalik dialect where the rule is to say, е. g. kv: ıliwane instead of 
ko-m iliwane ‘in the interior of the river’ (ku:k) or itt silata:ne instead 
of ifim silata‘ne ‘outside of the house’ (ite). 
Absolutive and Relative. 
S 36. — These names express a fundamental differentiation in the 
logic of the Eskimo grammar. When I am to give here a definition 
for practical use, it is not my intention to fix any definite view of 
this point in the grammar, of which opinions still differ. The entire 
question of subject, object and possession in the morphology of this 
language demands a comprehensive investigation in connection with 
1 There is here an evident conformity to the Eskimo tongue on the Baffin island 
west of the Davis straits where the same fluctuation of the final sounds was 
observed by F. Boas (cf. my Eskimo sketch, 1911, pp. 985 and 987, note 1) but 
I cannot say that I have noticed in East Greenland any constant difference in 
the pronunciation of men and women in this respect, such as Boas asserts with 
regard to the Baffinlanders. 
