68 



R. H. MATHEWS. 



I had the subject sufficiently in hand by 1898 to enable 

 me to take the responsibility of communicating a short 

 article to the American Philosophical Society containing 

 the following table of the intermarrying laws of the 

 Arranda tribe, 1 in which I stated that descent was through 

 the females. 



Table I. 



Cycle. 



Wife. 



Husband. 



Offspring- 





/Purula 



Pananka 



Bangata 



A 



Ngala 



Knuraia 



Paltara 





Bangata 



Mbitjana 



Ngala 





^Paltara 



Kamara 



Purula 





' Pananka 

 Knuraia 



Purula 



Kamara 



B- 



Ngala 



Mbitjana 





Kamara 



Paltara 



Knuraia 





Mbitjana 



Bangata 



Pananka 



The four women in the column headed "Wife" in the 

 upper half of the table constitute the Cycle A, and have 

 perpetual succession amongst themselves. Purula has a 

 daughter Bangata, who has a daughter Ngala, who has a 

 daughter Paltara, who has a daughter Purula, thus revert- 

 ing to the commencement of the series. The lower half 

 of the table or Cycle B has exactly the same constitution. 

 This is sufficient evidence that the devolution of the section 

 names is through the women. Every one of the daughters 

 would of course have brothers belonging to the same 

 section name as herself. 



Taking the first line in the table, we see that Pananka 

 marries a Purulu woman, who is his tabular or No. I wife, 

 and the offspring is Bangata. But Pananka could instead 

 marry a Ngala woman, as his alternative or No. II spouse. 

 Rev. L. Schulze 2 was the first to report the alternative 



1 Proc. Amer. Philos. Soc, xxxviii., p. 76 s table I. 



2 Trans. Eoy. Soc, S. Aust., xiv., 223 - 227. 



