OF THE HUMAN FAMILY. 411 



througli centuries of time, demonstrated by its preservation in such a number of 

 independent channels. 



3. Gujarathi. This system is also in full and and minute agreement with the 

 Hindi and Bengali, as will be seen by consulting the Tabic. It is chiefly interest- 

 ing as confirmatory of the truthfulness of the latter ; and for the additional testi- 

 mony which it furnishes of the stability of the system in its present condition. 

 The features in which it deviates from, as well as those in which it agrees with the 

 Tamilian are also constant in the Gujaratlii. 



4. Marathi. The same remarks that have been made with reference to the last 

 form are equally true of the Marathi. There are but two particulars in which 

 there is any noticeable diiference between the Marathi and those previously 

 explained. The first consists in the presence and use of special terms in the 

 ISIarathi system, for elder and younger brother, and for elder and younger sister, 

 which have before been considered ; and the other of the absence of the Polish 

 method of distinguishing the children of uncles and aunts. The failure to adopt 

 this method tends to confirm the inference of the Sanskritic origin of this method 

 of discrimination. For a further knowledge of this form, reference is made to the 

 Table. 



From the foregoing exposition of the Gaura system of relationship a definite 

 impression of its present characteristics has been obtained. The form which pre- 

 vails in the five remaining dialects must be ascertained and compared with those 

 given before the question of the true position of the Gaura system can be fully 

 determined. Presumptively the same form prevails in all of the dialects ; but at 

 the same time if the original system were the same as the Tamilian, other traces 

 than those akeady found may still exist in the unrepresented dialects. There are 

 two hypotheses, as before intimated, with reference to this system, each of which 

 has some basis of probability. First, that it is an independent variety of the 

 classificatory system, and has remained unchanged, in its radical features, since the 

 advent in India of the Sanskrit stock ; or, Second, that it was originally Turanian 

 of the Tamilian type, and has been modified to the extent of losing several of its 

 radical characteristics under the long-continued pressure of Sanskritic influence. 

 Upon the first hypothesis, in addition to what has previously been stated, it may 

 be remarked that it would exclude all influence from Sanskrit sources upon its 

 formation. If adopted, we must suppose that they voluntarily abandoned their 

 own descriptive system and accepted, in its place, the form of a barbarous people, 

 contenting themselves with the substitution of their own terms of relationship in 

 the place of the aboriginal. Upon the second, which is much the strongest 

 hj'iJothesis, it may be said, first of all, that the system is un-Sanskritic, and, therefore, 

 must have taken its origin without the Aryan family. Secondly, that when the 

 two peoples became united, two radically diff"erent systems of consanguinity were 

 brought into collision, and held in antagonistic relations until a new system was 

 constructed. Thirdly, that the resulting system would represent in the source of 

 its several parts the amount of influence each was able to impress upon it. Lastly, 

 That the Sanskritic influence would be directed with greater force against the 



