236 KINDRED GROUP-MARRIAGE 



For husband's brother we have a word running 

 through several Aryan tongues, and pointing to an 

 Aryan word daiver as its source. Thus Sanskrit devdr, 

 Greek Sarip (for Sa/^'p), Latin levir, Lithuanian deverls, 

 probably Armenian tagr, Anglo - Saxon tdeor, and 

 possibly O.H.G. zeihhur (this may = zuehir = suueher = 

 sweher, for it is glossed both socer and levir) are from 

 this source, and denote husband's brother. The root of 

 the word appears in Aryan daidi, denoting share, as in 

 Sanskrit ddyate, Greek Salco, whence &u?, a meal, and 

 English tide (time), German zeit, all marking a division 

 or share. 1 It would thus appear that the original sense 

 of devdr is the husband's brother as sharer. With all 

 the folklore and other evidence that w T e have for the 

 existence of the levirate custom, both before and after 

 the death of the husband, there can be little doubt that 

 it is the wife that the share refers to. 



Thus, whether we turn to the words for sister-in-law 

 or brother-in-law, we find primitive meanings strongly 

 supporting the hypothesis of kindred group-marriage, 

 and very hard to reconcile with an exogamic patriarchal 

 system. Further, the terms for son-in-law arid daughter- 

 occurs in the Irish legend of Cuchulainn, who was the son of King Lug from his 

 sister Dechtere, who shared his sleeping-apartment. The boy was reared by 

 one of the sisters of his mother and of the king, and the king remarked that 

 there was "little for her to choose between her own son and her own sister's 

 son" (see Rhys, Hibbert Lectures, 1886, p. 431). The German geschwister for 

 brothers, sisters, and cousins tells the same tale. I expect Norse systkin, 

 Danish SQsJcende, now used for brothers and sisters, originally included all the 

 children of sisters. In this respect the Danish s&skendeseng, a rough bed rigged 

 up on the floor for travelling companions, is perhaps not without suggestiveness 

 for the old group customs : see my remarks in Ashiepattle, p. 80. 



1 Here again, as in the mal root, I am inclined to think that the sharing 

 notion first led to a name for the common meal, and then that the notion of 

 time arose from the meal epoch : see p. 146. 



