STUDIES IN HUMAN HEREDITY 



305 



Mendelian expectation. Allowing, however, for quite a number of prob- 

 ably homozygous recessives, and occasional imperfection of dominance, 

 the proportion comes to accord fairly well. Perhaps an equally legitimate 

 assumption respecting the pedigrees in question would be to regard all 

 crosses as between RRs and DRs. The proportion should then be one to 

 one. Allowing then for a number of DR by DR crosses, the proportion 

 should approximate closely to the one obtained. 



Professor Hodge, of Winthrop College, South Carolina, has lately had 

 one hundred of my printed questionnaires filled out by left-handed students. 

 Charts, figures 22 to 35, give the data of the most complete of these. These 

 include numerous cases of nine-, ten- and eleven-child fraternities. In 

 this set of pedigrees we have the advantage of large childships. The total 

 number of children in these one hundred families is four hundred and 

 ninety-eight (498) ; of these eighty-nine (89) are left-handed (sixteen girls, 

 twenty-eight boys, and forty-five of whom the sex was not given). The 

 proportion of left-handed to right-handed is therefore as eighty-nine (89) 

 to four-hundred and nine (409), or as one to four and six-tenths. This is 

 slightl}^ greater than for the DR X DR cross, but is sufficiently close after 

 due allowance for the preponderance of large childships to be significant 

 as indicating the recessive character of left-handedness. 



U 



*6666ii 



^ 



■nrrr^ 



Fig. 22. 



S 



*4 



*3 Normal ciilldren of unknown sex. 

 :^4 Left-handed children of unknown s 



Fig. 23. 



•n-n •-Hn 



66666ii r 



Fig. 24. 



Fig. 25. 



