[The family and the sept.] 13. We have seen that it is in the groups or camps of the vagrant tribes like the Beriya, Hâbûra and Sânsiya, that we must look to find what is perhaps the most primitive form of human association, and that the family was almost certainly not the primitive unit, but the sept. The family, in short, arose out of the sept when the stage arrived at which paternity and the incidents connected with it came to be recognised. But of the real tribal form of caste in which the association is based on actual or assumed community of blood through a common ancestor, we find little or no trace, except as Mr. Ibbetson [30] showed to be the case among the Pathâns and Bilûches of the western frontier, who are foreigners in this part of India. But even here the fiction of common descent is being gradually weakened by the wholesale admission of outsiders into the fraternity, who do not even pretend to be able to establish a genealogical connection with the original founder of the sept. Here, too, the differentiation of industries is leading to a distinction, even among the members of the association linked together in theory by the bond of blood. In theory any Pathân, Mughal or Sayyid may marry any girl of his tribe; but if he falls in social position or adopts any degrading occupation his difficulty in marrying into a respectable family is as difficult as it would be in Germany or even in some grades of English society for a parvenu to marry into a family whose claims to rank are undisputed.