586 



NATURE 



\April 2\, 1 88 1 



different from the Malayan — his second order of terms 

 which he has named the Turanian system of relationships. 

 He regards the terms in this system also as accurately 

 describing, " as near as the parentage of children could be 

 known," the relationships existing at the time when they 

 came into use. It differs from the Malayan in including 

 words for cousin, uncle and aunt, and nephew and niece — or 

 words which Mr. Morgan has so translated. It will be found, 

 however, that Mr. Morgan does not use the punaluan family 

 in accounting for any one of the Turanian terms. Those of 

 them which coincide, or parily coincide, with the Malayan 

 terms he had already accounted for by the hypothesis of the 

 consanguine family, and he does this over again ; the ethers 

 he accounts for, or tries to account for("Ancient Society," 

 pp. 442-445), by means of exogamy alone. His reasoning 

 is exactly what it would have been had the punaluan 

 family never occurred to him. Indeed it has been an 

 emban-assment to him ; he has had to keep it out of his 

 reasonings. For the punaluan family is, ex hypothesi, in 

 two forms, and neither form could, "as near as the parent- 

 age of children could be known," yield both the Turanian 

 sense of father and the Turanian sense of mother. Where 

 the husbands were punalua, Mr. Morgan's reasoning 

 would make them all, though not brothers, fathers ol" 

 children born within the group, and it would exclude 

 their brothers from being considered fathers. But, in 

 the Turanian system, a father's brothers are called 

 fathers. Similarly where the wives were punalua, Mr. 

 Morgan's reasoning would make them, though not sis- 

 ters, all mothers of the children of each of them, and 

 would exclude their sisters from being considered as 

 mothers. But, in the Turanian system, a mother's sisters 

 are called mothers. Mr. Morgan has not failed to sec 

 this, and he has actually again framed a subsidiary 

 hypothesis to give his hypothesis of the punaluan 

 family a chance of living. This is (see "Ancient 

 Society," p. 445) that where a group of sisters married 

 men who were not brothers, they also became the wi\es 

 of all the brothers "own and collateral" — that is, all 

 the brothers and one-half of the cousins, however far 

 removed— of each of their husbands ; and, similarly, 

 that when a group of brothers married women who were 

 not their sisters, they also became the husbands of all the 

 sisters and one-half of the cousins of each of their wives. 

 All that need be said of this subsidiary hypothesis is that 

 it gives quite a new look to the punaluan family — and 

 that the effect of it, like that of the secondary hypothesis 

 formerly noticed, is to deny us all chance of judging 

 whether the principal hypothesis is a good or a bad one. 

 The justification offered for it is that "the system (the 

 Turanian) treats all brothers as the husbands of each 

 other's wives, and all sisters as the wives of each other's 

 husbands, and as intermarried in a group" — but that is 

 equivalent to saying that the system has taken no impres- 

 sion of the punaluan family, and gives no countenance to 

 Mr. Morgan's hypothesis. As, apart from " the system," 

 he finds nothing to say for it, it is difficu't to see how 

 any one can resist the conclusion that that hypothesis 

 must be dismissed, and that it must be ranked among the 

 wildest chimeras that have ever possessed thebrain of a 

 man of science. 



Now, do Mr. Fison and Mr. Howitt give in any degree 

 to Mr. Morgan's hypotheses the support of which they 

 are in need? The answer must be no — and must be no 

 even if we receive as facts the assumptions as to fact 

 from which they set out. Mr. Howitt accepts both the 

 consanguine family and the punaluan family, while Mr. 

 Fison offers himself as the advocate of the latter only. 

 But Mr. Howitt has nothing new to say for the con- 

 sanguine family ; he believes in it, and argues from it as 

 if it were known historical fact — that is all ; and so of it 

 no more need be said. What then do his colleague and 

 he find to say for the punaluan family .' Literally, not a 

 word. Mr. Howitt simply takes it for granted as he does 



the consanguine family. Mr. Fison, in beginning, under- 

 takes to show that it results logically fro a his hypothesis 

 — for it is no more than that — of "exogamous inter- 

 marrying divisions." but he does not attempt to do so. 

 And, in fact, his "intermarrying divisions" are quite 

 different from the punaluan family, and leave no need for 

 it, and no room for it ; that is, his hypothesis is different 

 from and exclusive of Mr. Morgan's. In Mr. Fison's 

 hypothesis, a group of men who are considered brothers 

 and a group of women who are considered sisters — being 

 the men and women of the same generation in two 

 divisions which intermarry with each other, and only 

 with each other — are by birth husbands and wives to 

 each other ; whereas, in the punaluan family, when the 

 husbands are brothers the wives are not sisters — they are 

 punalua ; and when the wives are sisters the husbands 

 are not brothers— they are punalua. Men who are 

 brothers are restricted to women who are each other's 

 sisters, on Mr. Fison's hypothesis ; but, on Mr. Morgan's, 

 men who are brothers marry women who, as a rule, are 

 not each other's sisters. The marriage law shown in 

 Mr. Fison's hypothesis would have to be given up before 

 the punaluan family could have a chance of issuing out 

 of the intermarrying divisions. Then, as Mr. Fison 

 justly observes, his intermarrying divisions " would have 

 precisely the reformatory effe.t " which Mr. Morgan 

 attributes to the punaluan family — so that, given the 

 divisions, the punaluan family would not be needed for 

 reformatory purposes ; and as Mr. Fison's view is that 

 the totem clan grew up within his divisions, while their 

 marriage law still subsisted, the punaluan family would 

 not be needed to give birth to the clan (which Mr. Morgan 

 says it has done). And, clearly, there would be no more 

 room than need for it. It thus appears that, instead of 

 supporting the hypothesis of the punaluan family, Mr. 

 Fison has put it aside, and offers an improved hypothesis 

 (suggested, no doubt, by Mr. Morgan's) in place of it. 

 We have seen that he does not accept the consanguine 

 family either. He does not, indeed, repudiate it. But 

 to connect it with his intermarrying divisions seems to 

 him so diffi:ult that he thinks the one could have been 

 changed into the other only through the intervention of 

 " a higher power." He is not afraid of the ridicule to 

 which he might be exposed were he to account for the 

 first formation of the divisions by such a hypothesis ; but 

 he thinks it unnecessary to go bjhind them. We have 

 now shown in what manner Mr. Fison supports Mr. 

 Morgan — and we have shown that Mr. Morgan is in no 

 position to give any support or countenance to him. 



To show that the Turanian terms would result logically 

 from his own hypothesis is what Mr. Fison has attempted. 

 There are in a tribe two divisions which do not permit 

 marriage within the division, and are restricted to inter- 

 marrying with one another. All the men in one division 

 are the husbands of all the women of the same generation 

 in the other ; the wife does not come into the husband's 

 division ; and descent is reckoned through the mother. 

 The group of men marries the group of women ; and it is 

 the group that is husband, the group that is wife, the 

 group that is father, mother, son, or nephew ; every 

 person in it taking, however, all the relationships that 

 arise to it. Such is the hypothesis. Seeing that the 

 relationships are called group relationships, it might be 

 thought that Mr. Fison considered the Turanian terms 

 to have been, in the first instance, something other than 

 terms of blood-relationship, say terms of address ; but he 

 denies that they are terms of address, and regards them 

 as having been real relationships from the first. 1 1 what 

 natural sense of relationship, however, a group — or the 

 women in it other than the actual mother — can be mother 

 of a child he does not tell us ; and till he can make this 

 plain, his theory must be held to be as untenable as the 

 hypothesis of the consanguine family. As for his demon- 

 strations (Q. E.D. at the end of each) of the Turanian 



