132 ETHNOGRAPHY. 



husband, rvahine, wife, and tamaiti, child, are introduced after each 

 name in the respective columns; thus 



O Haloa te tane, O Hinamanouluae ta icahine, O Wai.a te tamaiti. 



O Waia te tane, O Huhune to, wahine, O Hinanalo te lamaiti, &c. 



This, it will be seen, makes of it a species of verse, with, in fact, a 

 greater approach to rhythm than most of the native poerns. Accord- 

 ingly, the recitation is made in a kind of chant, to a regular tune, and 

 any person who can retain in his memory a song of a hundred lines 

 can have no difficulty in remembering this genealogy. There is no 

 reason why it should not have been known to hundreds, in fact to 

 the whole priesthood of Hawaii, and any lapse of memory in one 

 would be corrected by the rest. The same observations will apply to 

 the genealogies preserved in the other groups of Polynesia. 



It is to be observed that this is not, properly speaking, a list of 

 kings, but merely of generations. In those cases, which frequently 

 happened, where two or more brothers succeeded one another on the 

 throne, their names are given in the column of children. Thus Liloa, 

 the eleventh in a direct line before Tamehameha, had two sons, Hatau 

 and Umi, of whom the first succeeded him, but was deposed for his 

 tyranny, and the kingdom transferred to Umi. Both these names, 

 with those of their respective mothers, are given in the genealogy, 

 but the former only among the children. For the same reason 

 Tirvalao and Talaniopuu, who immediately preceded Tamehameha, 

 are not given, because the line of descent is not traced through them, 

 but through the younger brother of the latter, Teoua ;* and his name 

 is therefore in the list, though he did not actually reign. These ex- 

 planations are necessary, because the number of years to be allowed 

 to a generation will be at least double that which we should assign to 

 a reign. Among a people like the Hawaiians, constantly engaged in 

 wars, in which the chiefs are expected to take an active part, the ave- 

 rage duration of a reign can hardly be estimated at more than fifteen 

 years, while there is no reason for assigning to a generation a shorter 

 period than that at which it is commonly rated, about thirty years. 

 The people do, indeed, marry younger than in more northern regions; 

 but this consideration is counterbalanced by the fact, which appears 

 from the genealogical table, that, in many instances, the pedigree has 



* This is a name which is given here in place of the unwieldy appellation of Tcdani- 

 tupuapaitalaninui, which appears in the genealogy. 



