482 AssiiJ-fafi're Jfatiiiff in Mau 



In the present study we endeavour to measure how far there is an assortative 

 mating betwcen husband and wife with regard to thcir durations of life, or rather 

 witli regard to the gencral pliysical characters on which this duratiou dopeiids. It 

 has been shewn iu Bioinetrika* that duration ot" life is au inherited character, and 

 heuce assnrtutivo luating iu respect to this characti-r niay probably bc lookod upon 

 as the mating ot' iiko Stocks. 



2. Our knowledge of assortative mating in uiau may be summed up as follows. 

 It was first noted whcii draliug with Francis Galton's Family Records. K. Pearson 

 höre füund a correlalion of •09 for statnre"f and 'lO for eye colour;!:. In Pcarson's 

 niore extensive series of Fauiil}- iMeasuivnients, the foilowiug value.s were fouud§: 



Assurtalive Mating in Man: lÜOU Cases. 



Stature .... -2804. 



Span 1989 



Left Forearm . . . -1977 



Mmn Valtte . . . •2257 



Fnrther, values between ^14 and 20 were found for cross assortative mating, 

 e.(/. between stature in husbaud and forearm in wife. 



Now tliis very scu.sible value of ^226 is either diie to, (i) a real couscious 

 or unconscious a.ssortative mating in man, nr (ii) tu individnal men and woinen 

 mating within sub-races, local races or other limited classes. 



Now the infiuence of (ii) can be more or less etfectively determiued iu our 

 present investigation by taking data within («) a fairly limited class and (b) within 

 limited districts. If auy .small local differences produce the observed result then 

 WC ought to find correlation sensibly as largc as the a.ssortative mating correlatiou, 

 when WC niake randoiu couples of men and womeu from the same locality and put 

 several localities into thi^ same table. 



In coUccting our material we were acconlingly guidcd by the desire to get a 

 fairly narrow distriet, and one if possible where there would be no unintentional 

 selection of husbands and wives living appro.xiniately to the same age. By using 

 peiligree tables we can avoid any selection of this kind, it' Imth husbaud and wife 

 are given, but iu many pedigreos the ages of the men only are giveu. The records 

 of the Society of Friends are an e.xceptiou to this rule, and that Society forms a 

 (airly huinogeneous sub-cla.ss of the Community: indeed so compact that attention 

 has more thau once been drawn to the ditierentiation of its actuarial coustants 

 from those of the comnuinity at large. 



Data from pedigi-ees being limited owing to the too frequent absence of the 

 dates of birth and death uf the wives, we not unnaturally turned to the records 

 provided by tombstones. In collecting such material, it is esseutial to work on 



* Vol. I. p. 50. M. Beeton aud K. Pearson: O/i the Inhritancf of the Duration of Life, cto. 



+ l'hil Tram., Vol. 187 .\, p. 273. 



+ Phil. Trans., Vol. l'.l.l A, ]). IIH. § Biomelrika, Vol. n., p. 373. 



