.•September 29, 1911] 



SCIENCE 



391 



concept of mana as underlying primitive 

 magic. Here, again, I do not wish to deny 

 that a concept such as that denoted by 

 mana may be a primitive element of magic ; 

 all that I wish to point out is that the 

 Melanesian evidence can not properly be 

 used to support this view, for the use of 

 the term in connection with magic in 

 Melanesia is not primitive, but secondary 

 and relatively late. 



The point, then, on which I wish to insist 

 is that if cultures are complex, their analy- 

 sis is a preliminary step which is neces- 

 sary if speculations concerning the evolu- 

 tion of human society, its beliefs and prac- 

 tises, are to rest on a firm foundation. 



I have so far dealt only with Melanesia. 

 It is obvious that the same principle that 

 analysis of culture must precede specula- 

 tions concerning the evolution of institu- 

 tions is of wider application, but I have 

 time only to deal, and that very briefly, 

 with one other region. 



No part of the world has attracted more 

 attention in recent anthropological specu- 

 lation than Australia, and at the bottom 

 of these speculations, at any rate in this 

 country, there has usually been the idea, 

 openly expressed or implicitly understood, 

 that in the culture of this region we have 

 a homogeneous example of primitive human 

 society. From the time that I first became 

 acquainted with Australian sociology I 

 have wondered at the complacency with 

 which certain features of Australian social 

 organization have been regarded, and espe- 

 cially the combination of the dual organi- 

 zation and matrimonial classes with what 

 appear to be totemie clans like those of 

 other parts of the world. This coexistence 

 of two different forms of social organiza- 

 tion side by side has seemed to me the 

 fundamental problem of Australian so- 

 ciety, and I confess that till lately, ob- 

 sessed as I see now I have been by a crude 



evolutionary point of view, the condition 

 has seemed an absolute mystery." A com- 

 parison, however, of Australia and Mela- 

 nesia has now led me to see that probably 

 we have in Australia, not merely another 

 example of mixture of cultures, but even 

 another resultant of mixture of the same or 

 closely similar components as those which 

 have peopled Melanesia, viz., a mixture of 

 a people possessing the dual organization 

 and matrilineal descent with one organized 

 in totemie clans, possessing either patri- 

 lineal descent, or at any rate clear recog- 

 nition of the relation between father and 

 child. This is no new view, having been 

 already advanced, though in a different 

 form, by Graebner^^ and P. W. Schmidt.^^ 

 If further research should show Australian 

 society to possess such complexity, it will 

 at once become obvious that here also eth- 

 nological analysis must precede any theo- 

 retical use of the facts of Australian so- 

 ciety in support of evolutionary specula- 

 tions. 



It may be objected that we all recognize 

 the complexity of culture, and indeed in 

 the study of regions such as the Mediter- 

 ranean, where we possess historical evi- 

 dence, it is this complexity which forms the 

 chief subject of discussion. Further, 

 where we possess historical evidence, as in 

 the cases of the Hindu and Mohammedan 

 invasions into the Malay Archipelago, all 

 anthropologists are fully alive to the com- 

 plexities and difficulties introduced thereby 



•"I may note here that Mr. Lang, after having 

 considered this problem from the purely evolu- 

 tionary standpoint ("Anthropological Essays pre- 

 sented to E. B. Tylor," p. 203), concludes with 

 the words, ' ' We seem lost in a wilderness of diifi- 

 eulties. ' ' 



»Zeitsch. f. Ethnol, 1905, XXXVII., 28, and 

 "Zur australischen Religionsgeschichte, " Globus, 

 1909, XCVI., 341. 



"See especially Zeitsch. f. Ethnol., 1909, XLL, 

 340. 



