44 SYSTEMS OF CONSANGUINITY AND AFFINITY 



method of designating collateral kindred, which is the most important part of the 

 system, is wanting. It is for this reason of but little value for comparison. Since 

 both the Lithuanian and Lettish dialects are still spoken, the system of relationship 

 of each of these nations is still a living form. The absence of the Lithuanian, 

 therefore, is the more to be regretted, since it might have shown the original 

 Slavonic form, and thus tended to explain its peculiar features. 



VI. Celtic Nations. 1. Erse. 2. Gaelic. 3. Manx. 4. Welsh. 



1. Erse. — The forms in the Gaelic and Manx are in so near agreement with the 

 Erse that they will be considered together ; but the illustrations will be taken from 

 the latter. 



The Celtic system, as it appears in the forms of these three nations, is purely 

 descriptive. It is more strictly the typical form of the Aryan family than the 

 Iloman, and on some accounts should have been first presented. But as the Roman 

 was based upon the same original, and embodies all the developments from it sub- 

 sequently made, it furnished a better starting-point for the exposition of tlie 

 descriptive system. Whilst the Turanian and American Indian systems employ 

 special terms for every recognized relationship, and are therefore non-descriptive, 

 the Celtic, possessing no special terms except the primary, is descriptive, pure and 

 simple ; and thus holds the opposite extreme. The difference, as will appear in 

 the sequel, is fundamental. There is every probability that the Erse and Gaelic 

 forms have remained as they now are from a very early period. 



AVhere relatives by blood and marriage are described, without exception, by a 

 combination of the primary terms, it might be supposed to indicate the absence of 

 any positive system of relationship; but this would be an erroneous inference. 

 Such a form is essentially affirmative. To describe kindred in this manner we 

 must ascend step by step, by the chain of consanguinity, from Ego to the common 

 ancestor, and then descend in the same definite manner in each collateral Hue to 

 the particular person whose relationship is sought ; or, we must reverse the process, 

 and ascend from this person to the common ancestor, and then down to Ego. By 

 this means the natural outflow of the generations is recognized, the several colla- 

 teral lines are preserved distinct from each other and divergent from the lineal, and 

 absolute precision in the description of kindred is reached. So far it contains a 

 positive element. In the second place, to resist for ages the invention or adoption 

 of special terms for the near collateral relationships which are so constantly needed 

 in domestic life, evmces a decisive, not to say pertinacious, preference for the 

 descriptive method. Although this form suggests from within itself a certain num- 

 ber of generalizations of kindred into classes, with the use of special terms for these 

 relationships in the concrete, yet a system must be developed up to and beyond the 

 Roman standard form to render the use of these common terms definitely expres- 

 sive ; or, in other words, to secure the precision of the purely descriptive method. 

 As a domestic institution the system necessarily possesses the elements of perma- 

 nence ; and its modifications are the slow products of time and growth. Beside 

 the adoption of the Roman as our legal form, the only changes in the English sys- 

 tem within the last five centuries, so far as the writer is aware, is the restriction 

 of the terms nejijievy and niece to the children of the brother and sister of Ego, and 



