0. A. Merritt Hawkes 271 



Consideration must be now given to the possibility that S is some- 

 times a heterozygous form. The six matings (S' x (A or B) produced 

 15 X and 9 of the other types, a very large proportion of L, unless S is 

 sometimes heterozygous or unless the L factor is what may be described 

 as an excessive dominant, and the matings L x *S' by no means indicate 

 that. In this connection the family in Fig. 6, Tree H, (i), is interesting, 

 where the mating >S' $ x J. J" produced two male L. Now this female 

 S may be a heterozygote, being the child of the mating L ^ y. S $ , or 

 the two male L of generation three may themseh'es be heterozygotes, a 

 point which can only be settled during the next decade. This may be 

 a case in which the heterozygote appears as an S because it is a 

 female. 



There is certainly a sex factor in the S type and this may be 

 accounted for by the existence of certain female heterozygotes appearing 

 as 8 type ; on the contrary, there may be a tendency for the male 

 heterozygote to appear as an L. It is possible that a comparison of the 

 matings L ^ with A, B or S, and L $ with A, B ov S, may throw some 

 light on the sex difference. Of the former, there are 34 matings, 

 producing 72 L, 14 >S', 13 A and B, or 72 L type to 27 of the other types, 

 a proportion of 16:6. Of the latter (L $ with A, B or ,Sf) there are 22 

 matings producing 70 L, 9 »S', 10 A and B, or 70 L to 19 other types 

 or approximately 23 : 6. These results suggest at once that more male 

 than female L are heterozygotes. Reference must here be made to Fig. 6, 

 Tree G, in which it is tlie female L which appears heterozygous rather 

 than the male L. 



The above results must be controlled by an enquiry inti > the proportion 

 of the sexes in the groups of offspring arising from the above two groujjs 

 of matings. In both cases the A and B types will not be counted as 

 they show no sex difference. 



The mating L J' x (A, B m- S) pi'oduces 86 // and .S' ciffspring, 

 divided thus: L, 35 </ and 37 ? ; ^^ 3 ^ and 11 ? , i... </ : ? " ^ : 4, 

 and there are 19'4 per cent, of >S' type. 



The mating L $ x (^1, £ or S) produces 79 L and *S' offspring divided 

 thus: L, 29 J' and 41 ? ; S, 4 cf and 5 ?, i.e. </:?:: 3 : 4, and there 

 are 12'8 per cent, of 6' type. 



The sexes in both cases are approximately the same, but a larger 

 proportion of S is produced where the L parent is male rather than 

 female. Thus, the above deduction is confirmed, that when the L type 

 is heterozygous, it is more frequently male than female. 



