I do not, therefore, agree with the following remark of Sir Laurence Gomme (_op. cit._, p. 232): "It is of no use preparing a genealogical tree on the basis of civilized knowledge of genealogy if such a document is beyond the ken of the people to whom it relates. The information for it may be correctly collected, but if the whole structure is not within the compass of savage thought it is a misleading anthropological document." If it is possible at all to collect a genealogy, that means that individual kinship exists in such a community; in other words the "structure is within the compass of savage thought," only it is not apprehended by them in the same manner as by us. It is certainly true that in many cases the knowledge of this aboriginal apprehension is essentially needful for a sociologist. This has been argued in the text. [575] _Camb. Univ. Exp._, v. chap. iii. on Kinship, pp. 129 _sqq._ In particular, pp. 142-152, under the headings "The Functions of certain Kin," and "Kinship Taboos." [576] Recorded by Dr. Haddon, _loc. cit._, v. p. 210. Messrs. Fison and Howitt in their treatise on Australian kinship[577] do not give anywhere a clear definition of the concept in question. The only place where something like definition is given is page 121, where kinship is said to be "membership in the same tribal division," and where there is an acknowledgment that beyond "kinship" there still lies "personal relationship" between the parent and child. This is true, but this is only the first distinction upon which the actual discussion of the problem ought to be based. That the want of such a discussion is a serious defect in the book is obvious. [577] _Kam. and Kurn._