Axsortative Mafiiif/ in Man 487 



luiwcvcr, liarilly aiiy sioiiilicaul dill'. reticc hetwccii llu- two coiTclatiuns ; wo have 



l'ii' 'a n- 



lu iiiariird |iair.s /'a ,j-= - 0»)2 I .") + 0-()139 



In randdiii <'(MipI,-.s j-a'»- = - 0-G576 + 0-0128 



\\\r (HHeiviice betwL'L'ii tlie two beinj^- practically ncglitrible. 



It appears, then, that thc currflatiun hct.wccii tlu' leiigth of lifo in husband and 

 in wifc is not sensibly due to any of tlic possibie environmoiital cffucts considerud. 

 It niust thereforo bo eitbcr a roal phcnoincnon, or a rcsnlt of sonie süiection in the 

 records iVnin which the data witc nblaincd, a pussibilit.y whicli iias already bcen 

 dcalt with and disniissi'd. 



(7) Wc arc! thus forced to conclnde that tho convlation actuaily observcd 

 bctween the iength of life in two marriod pensons is a nieasure of a real tendency 

 tdwards huniogainy, comparable with other cases of assortative mating. CoUecting 

 the resnits obtained, we find the foUnwing vahies. 



Currelation between Length of Life in llnHhnnd iiiid Wife. 



Wensleydale and Di.strict ... 0-2:iUU± 0-0244 



Oxfordshire O-2ä()O±n-0211 



Society of Fricnds 0-1999 + 0-0l>12 



Mean 0-2-233 



These vahiea never differ by twice the probable error of their differenccs, and 

 they are in remarkably good agreement with Pearson's value for the correlation of 

 othcr physical characters in husband and in wife, the mean in which is 0'2257, thus 

 affording further evidence for the propositions (1) that a very sensible aniount of 

 homogainy does occur in man, and (2) that length of life is a character which is 

 subject to selection, as it has already beeu showu to be capable of hereditary 

 transmission*. Both these j)ropositions are clearly of great importance to the 

 practical student of human evohition. 



\Ve thus reaeh the conceptii.ai that husband and wife are as much alike as 

 uucle and nieee, and probably as much alike as, if not more alike than, tirst 

 Cousins; this is not only true for definite physical organs like stature and forearm, 

 but also for the general physical Constitution. Such a dogree of resemblance is 

 one which could certainly not have been auticipated, and which ma}' even appear 

 paradoxical to mauy. 



The London cemeteries present us with an even higher degree of correlation 

 between the durations of life of husband and wife,namely 0'4'204! + '0176. But we 

 believo, as we have indicated above, that this apparent doubling of the assortative 

 matiug is solely due to the trausitory nature of the population, — a spurious corre- 

 lation produced by husband and wife being less frecpiently interreil in the same 

 grave whcn they die at a long interval apart. 



* M. Btcton and.K. Pcarsou : liuij. Hoc. l'roc. Vol. 05, pp. 2'JO — 305. 



G2— 2 



