14 TORRE Y A . 



of the plant. There can be no serious doubt that the sex-expression of other 

 angiosperms, many of whose mutations parallel those observed in maize, 

 is likewise dependent upon the activity of many genes. 



A large proportion of the genes concerned in the sex-expression of con- 

 temporary species may have come down unchanged or little changed from 

 primitive angiosperms. Those early species were themselves the outcome 

 of a long evolutionary history in whose course had been developed a complex 

 genotype. As has been seen, the passage from hermaphroditism to other 

 sexual conditions need involve only a comparatively small number of muta- 

 tions of genes already present. One step — perhaps in general the final step 

 if maize may serve as an example — involved the establishment in some mem- 

 bers of heterozygosis with reference to one pair of genes as to which other 

 members of the species are homozygous. The pair of chromosomes bearing 

 this allelic pair now plays a part in sex-determination. 



Again to judge from maize, the selection of different mutant genes may 

 in different cases give rise to the same phenotypic result — namely, dioecism. 

 This example shows, too, how different pairs of chromosomes may in dif- 

 ferent cases come to function in sex-determination, as the X-Y pair appears 

 to function in some seventy-odd species of dioecious angiosperms. The rela- 

 tion of the sex chromosomes to the differentiating genes may vary from 

 species to species. In Rumex (13), as in Drosphila, the Y chromosome plays 

 no demonstrable part in sex-differentiation. In Lychnis, on the other hand 

 (17), its role is a positive one. In Fragaria, as in one of Emerson's derived 

 races, the "X-Y" pair characterizes the female of the species ; in all other 

 well-known cases in angiosperms, this pair is the property of the male. While 

 a partial picture is thus presented of the functioning of a pair of chromosomes 

 in sex-separation, no satisfactor}- explanation is yet available for the frequent 

 visible differentiation between the members of this pair. At the same time, 

 it is shown in more than forty investigated angiosperms that there is no 

 necessary correlation between the final genie differentiation which in a 

 dioecist influences sex and a perceptible difference in chromosome size or 

 appearance. 



The mutations that have determined the transition from hermaphroditism 

 have not produced in most cases, if in any, an absolute fixity of sexual char- 

 acter. Instead, whatever the inhibiting tendencies of a particular mutation, 

 it remains possible, under favoring conditions, for some or all of the old 

 potentialities to be manifested ; as when bisexual flowers appear on a monoecist 

 or dioecist. Xo rigidity of sex-separation seems to have been reached by the 

 angiosperm sporophyte such as characterizes the angiosperm gametophyte 

 or the gametophyte of a dioecious bryophyte. 



The conception which emerges of the genetic basis for sex can not be 

 satisfactorih' formulated in terms of so manv genes for maleness and so 



