x] Sex-limited Inheritance 177 



properties like those elsewhere proved to attach to other 

 dominants. 



One very curious observation made In the case of 

 grossulariata is that which remains to be stated. It was 

 communicated by Doncaster to the Dublin Meetlnc,^ of the 

 British Association (Sept. 1908). The experiments enu- 

 merated were, it will be observed, Incomplete in so far as 

 the mating \^'Adi gross, $ x lact. $ had not been made. The 

 results of this mating are now known. FamiHes thus bred 

 consist of mdAQS, grossulariata, ■Siwd females lacticolor\ In 

 other words, the ordinary wild grossulariata even in districts 

 where lacticolor is unknown, are in reality a race of which 

 the males are i^uvq grossulariata, though the females are in 

 reality hybrids of lacticolor, and so continue from generation 

 to generation. The normal female grossulariata and the 

 F^% grossulariata bred from lact. $ y. gross. $ are thus seen 

 to be Identical in composition. Whether a gross. $ has a 

 gross. ? for a mother, or a lact. $ for a mother makes no 

 difference to its composition and properties. This fact is 

 one of the most striking to which genetic research has yet 

 led. It affords strong confirmation of the Interpretation of 

 the series of phenomena given in the text, and enables us 

 to see in the evidence both as to the grossulariata case 

 and as to the other cases which follow, a consistent mass of 

 testimony all pointing in the same direction. In a recent 

 paper (114) Doncaster accepts the view here suggested. 

 It must of course still be remembered that attractive though 

 the present suggestion is, by reason of Its simplicity, we have 

 no proof that the natural scheme may not be more complex. 



The two cases next to be considered resemble that 

 of grossttlariata in the fact that reciprocal crosses be- 

 tween pure types give dissimilar results. Though both 

 examples to be discussed are only imperfectly explored, the 

 facts elicited are so curious that some preliminary notice of 

 them is called for. We may confidently anticipate that 

 further search will discover other comparable instances. It 

 will be seen that these phenomena point very plainly in the 

 same direction as those previously described, to the con- 

 clusion, namely, that, in the types used, femaleness depends 

 on the presence of a definite dominant iactor. 



p>. H, 12 



