212 R. H. MATHEWS. 



child, the result would have been the same. Such a mar- 

 riage is 'regular' or 'direct/ as in Table No. I., and could 

 not be disturbed by the totemic regulations, because the 

 parties could not possibly belong to the same totem. 



Instead of tracing the descent through Ippai's daughter, 

 Matha A, as above, let us suppose that Ippai's son, Murri A, 

 has married some strange Butha and produced a son, Ippai 

 E, then such a son could marry Ippatha D of the last para- 

 graph. This would be a brother's son's child marrying a 

 sister's daughter's child ; or a brother's daughter's child 

 marrying a sister's son's child. This constitutes the 'rare' 

 marriage spoken of in a previous page. In this case it 

 would be possible, in certain circumstances, for the totems 

 of the parties to be the same, and therefore the union 

 would not be sanctioned. 



Ippai's grandson, Ippai E, has still another source of 

 getting a wife. Ippai's maternal nncle, Kumbo, has a 

 daughter Kubbitha, who marries, and produces Matha F, 

 (not shown on diagram). Then Ippai E can marry Matha F. 

 The brother of Matha F could in like manner marry the 

 sister of Ippai E. That is, a man's grandchild marries his 

 uncle's grandchild. This is what I have called the 'alter- 

 native ' marriage ; and like the first, it is free from totemic 

 interference. 



A wife for Kubbi C could also be found by following the 

 descent of his mother's mother's relatives, or his mother's 

 father's relatives, or his father's mother's relatives, in an 

 analogous manner. Again, Kubbi O could marry a Matha 

 of the proper lineage. In other words, a man of any given 

 division can marry into one or other of the three remaining 

 sections, and also into his own. 



In the above examples I have traced the genealogy 

 through the grandfather, Ippai Al, down to Kubbi O, and 



