Both Yaos and Anyanja trace descent through the mother, and cannot marry within their own clan, which is, of course, the mother’s. The Yao clans are still clearly known and named. Mr. R. S. Hynde says: ‘The Yaos are divided among themselves into sub-tribes, stocks, or totemistic clans,[30] each with its own distinctive name, _e.g._ the Amwale, the Asomba, the Apiri clan. If you question them on the subject, they will usually be able to tell you this clan name, unless the person questioned be a slave, who, from various causes, may not know it.’ _Somba_ means ‘a fish,’ and _mwale_ ‘a girl’; _piri_ in Nyanja is ‘a mountain,’ though I have been unable to discover its significance as a Yao word. Thus, if Mwepeta, of the _Somba_ clan, marries Ndiagani, of the _Mwale_, their children will be _Mwale_, and none of them can marry a _Mwale_. Their nearest relation and natural guardian will be their mother’s brother, who (if the grandfather is dead) will be head of the family. But his children will be no relatives of theirs, as they belong to their mother’s clan, and she (by the rule) cannot be a _Mwale_. Any of Mwepeta’s sons or daughters will therefore be free to marry any of these cousins; but they could not marry the children of Ndiagani’s sisters, who would be _Mwale_. Some tribes of Anyanja east of the Lake have, in addition to this, a system of agnatic descent, through the father, called _chilawa_. This may be borrowed from the Zulus (who always count descent in this way, though the importance assigned to the maternal uncle is probably a survival of a former state of things), and the more southern tribes call it ‘the Angoni system.’ These last do not appear to have surnames; but those who recognise _chilawa_ do; these names descend in the male line, and show at once to what family a man belongs on his father’s side. ‘Although a person’s surname is not generally known to those who are not his near relations or intimate acquaintances, because, so to speak, he does not make personal use of it, and is not called by it, yet every one knows his own surname, and is ready to give it at once, if asked for it.’ The late Bishop Maples, who put the above facts on record, was of opinion that, in spite of the close relationship existing between the sister’s son and the mother’s brother, the father is really the head of the Nyanja family, and arrived at the conclusion that a distinction must be made between ‘kin’ and ‘blood’: ‘the mother preserves to her offspring the tie of kinship, the father that of blood.’