282 Shall. 



III. Other crosses bearing upon the relation of the gene IB 



to the gene F. 



It is now a familiar idea, first suggested by DE VRIES (1903, p. 504) 

 and frequently mentioned by recent writers, that a mutation which takes 

 place in the preparation of any gamete may be expected to produce a 

 heterozygous mutant, because of the practically prohibitive chances 

 against the meeting of two gametes which have independently under- 

 gone the same mutation. Those supposed mutants which breed true from 

 the first, and especially those which possess a new dominant character, 

 would seem to require some other interpretation, unless, as appears to 

 be true to a large extent in the Oenotheras, a hereditary mechanism 

 is involved by which heterozygotes may breed true. 



If the mutant is a Mendelian recessive, as in the case of L. dioica 

 angustifolia , it may be supposed to make its appearance only when a 

 union takes place between a sperm and an egg, both of which lack the 

 determiner for the dominant character of the parent type. The question 

 presents itself, therefore, as to the condition of the broad-leaf deter- 

 miner, B, in the strain which gave rise to the original angustifolia 

 mutant. From the results of the crosses recorded above in Tables H 

 and IY, it is clear that the females of normal broad-leafed strains are 

 homozygous for both the female gene F and its coupled broad-leaf gene 

 B. The broad-leafed males in the FB families, as we have also seen in 

 Table III, were heterozygous for both of the genes B and F, but what 

 is the condition of B in the males of normal broad -leafed strains? It 

 might be either homozygous, XBF.XBf, or heterozygous, XBF.Xbf. 

 If homozygous, the occurrence of a narrow -leafed male mutant must 

 have required a double mutation, affecting both an egg and a sperm. 

 If on the other hand, the males of the normal strains, generally, are 

 simultaneously heterozygous for the genes B and F, a single mutation 

 which removed the B from its combination with the F might give rise 

 to a narrow-leafed male, for in every cross half of the available sperms 

 (Xbf) would already lack the gene B, and would thus be ready to unite 

 with any XbF egg which might be presented. Prior to the presentation 

 of such a mutated egg, the existence of Xbf sperms could not be dis- 

 covered because they must always fertilize eggs of the type XBF. 



Now that heterozygous females are available, half of whose eggs 

 are of the type XbF, it will be easy to test the males and hermaphro- 

 dites of all the various normal strains with regard to the homozygous 



