OF TlIK HUMAN FAMILY. 37 



rural populations in Denmark, Norway, and Germany as well, still adli(>rc to the 

 native term. 



The first collateral line male gives the series, hro/hrr, hro/Jii-r's son, and hro/Jicr\'i 

 grand-child ; the second, /((7Z!t'/''s hrothcr, cousin, and cousin's grand-child ; and the 

 third, far-father's brother, father's cousin, father's cousin's son, and father's cousin's 

 grand-child. These illustrations reveal the character of the system. 



7. Swedish. — The Swedish form agrees so closely with the Danish and Norwegian 

 that it docs not require a separate notice. 



8. Icelandic. — The insulation of the Icelandic Teutons would tend to preserve 

 their form of consanguinity free from foreign influence. It has original terms for 

 grandfather and grandmother in afi and amma, and a term nefi for nephew, which 

 is given in the jNlithridatcs, but docs not appear in the Table. It has terms, also, 

 for first and second cousin, which are used concurrently with the descriptive 

 phrases. In form and method, howoAcr, it approaches nearer to a purely descriptive 

 system than any yet presented. 



In the first collateral line, male, the series is as follows : Brother, son of 

 brother, son of S07i of brother, and son of son of son of brother. It agrees with 

 the Celtic ia commencing the description at the opposite extreme from Ego, which, 

 although it may be an idiomatic peculiarity, is yet significant, and will reappear in 

 the Armenian and also in the Arabic. For the second collateral we have fatlter's 

 brother, son of father's brother, son of son of father's brother, and son of son of 

 son of father's brother. The same form, which is seen to be purely descriptive, 

 runs through the several lines. It follows strictly the natural streams of descent, 

 and makes each relationship specific. This realizes Avhat we understand by a 

 descriptive system. It is evidently nearer the primitive form of the Aryan family 

 than that of any other nation of the Teutonic branch. The advances made by 

 some of the nations, which it is the object of this comparison to trace, are seen 

 to be explainable. They have not proceeded far enough to obscure the original 

 form with which they severally commenced.^ 



» Nomonelatares of relationship develop from the centre outward, or from the near to the more 

 remote degrees. The primary terms would be first invented since we cannot conceive of any people 

 living without them ; but when the nomenclature had been carried to this point it might remain 

 stationary for an indefinite period of time. The Celtic never passed beyond this stage. By means 

 of these terms consanguinei, near and remote, can be described, which answered the main end of a 

 nomenclature. Further progress, 'or the development of secondary terras, would result from a desire 

 to avoid descriptive phrases. The first of these reached would, probably, be nepos, as elsewhere 

 stated, and made to include several classes of persons. Next to this would, probably, be terms 

 for grandfather and grandmother. In the Romaic, Hellenic, and Slavonic stock languages there are 

 terms for these relationships, which, it is somewhat remarkable, are distinct and independent of each 

 other. In the other dialects they are wanting. It would seem to follow that no terms for these 

 relationships existed in the primitive speech, and that the persons were described as "fatlier's 



father," and so on. 



Next in order, apparently, stand the relationships of uncle and aunt. Those do not appear to 

 hate been discriminated, in the concrete, in the primitive speech. A common term for paternal 

 uncle is foimd in the Sanskrit pairo]/o, Greek jMtros, and Latin 2Mtruus; but this term seems to be 



