718 PROCEEDINGS OF SECTION G. 



tlieless it is absolutely necessary to the theory, which is that 

 exogamy forbids marriage within the community, and that 

 the only way of getting a wife is by capture from another 

 community. If, therefore, any girls on either side are left 

 uncaptured, what is to be done with them ? 



3. In the third place, as a matter of fact, savages do not 

 practise female infanticide — that is to say, they do not kill 

 female children rather than males. The Hindu, who has 

 succession through males, and who depends on his sons for 

 feeding him with sacrifices when he has taken up his abode 

 in the next world, and who has to give a dower with his 

 daughter, will kill a female infant rather than a male, if he 

 practise infanticide at all ; but not so the savage, whose 

 descent is through females, and who receives payment from 

 the young fellow who wants his daughter. 



4. And, in the fourth place, no tribe on the face of the 

 earth ever had polyandry — or polygynia either — as its system 

 of marriage. Cases of polyandry are to be found in plenty 

 of .tribes, and everybody knows that the other form of poly- 

 gamy — plurality of wives — is frequent enough ; but as a 

 complete system of marriage, \t is an arithmetical impossibility. 

 It requires a disparity in number between the sexes such as 

 never existed in any normal community. Polyandry, of 

 course, means that every woman has several husbands. 1 

 say ^^ every woman," because every woman must be reckoned. 

 Savage society has no old maids in it. Then if the average 

 number of husbands to each woman be the lowest possible to 

 constitute polyandry — that is two — the men must be twice as 

 numerous as the women, which again is absurd. 



The same argument is fatal to the other form of polygamy 

 as a general system of marriage in any community. It is 

 only the powerful chiefs, and the men rich enough to buy, 

 who have a plurality of wives, and not more than one apiece 

 falls to the lot of the commoners. Everyone who has lived 

 among savage tribes will confirm this statement. 



But further, the evidence for polyandry is not at all 

 satisfactory. Even the special case on which Mr, M'Lennan 

 mainly relies — that of the Nairs — can be shown to be not 

 polyandry at all, but group-marriage ; and the main object of 

 this paper is to examine the so-called Nair polyandry in the 

 light of a precisely similar Australian example. 



The account of the Nair custom, as given by Mr. 

 M'Lennan from various authorities, is that " among the Nairs 

 it is the custom for one woman to have attached to her two 



