JOHN W. GOWEN. 



female class in defect of its expected ratio. The agreement be- 

 tween the proportions of twins for these data and the proportions 

 of twins on the theory that they are distributed in accord with 

 those of human twins, where identical twins are known to exist, 

 shows a slightly poorer agreement (P - =.24) than does the theory 

 of i \2.\ i. The data of this paper clearly show a distribution of 

 1:2:1, indicating that identical twins are seldom produced in 

 cattle. 



Independent confirmation of this hypothesis may be obtained. 

 The American Jersey Cattle Club describes the animals registered. 

 From the Herd Books of this Association 749 twin females, 168 

 twin males and 207 male and female were taken for a comparison 

 of their color markings. These twins have been taken and their 

 white markings, tongue color and switch color recorded. The 

 by-laws of this Society in 1913 do not mention any requirement 

 to indicate that the female twin can not be registered before she 

 is proved a breeder. Besides these, 523 pairs of females, the off- 

 spring of the same sire and dam, have been tabulated for these 

 colors. These last were found from the Registry of Merit tables 

 in conjunction with the Herd Books. Table II. shows the dis- 

 tribution for these data. 



TABLE II. 



Table II. shows that the percentage of cows which are alike in 



