86 Species Hybrids of Digitalis 



more favourable to the expression of the characters of its own nucleus 

 (the female nucleus) than of those brought in by the male nucleus. 



But although such an explanation would fit the present case of 

 Digitalis, it is difficult to apply to the case of (Enothera recently investi- 

 gated by de Vries\ in which the reciprocals follow the male parent'. 



Even in (Enothera, however, although the reciprocals resemble the 

 male parent in respect to size and shape of leaves and flowers, &c., they 

 take after the female parent in some features : e.g. {(E. biennis x (E. 

 muricata) has the green leaves of biennis and not the blue-green of 

 muricata. 



De Vries found that both reciprocals between the above two species 

 bred true for an indefinite number of generations. He has found also 

 that the result of crossing the hybrids together is different according to 

 which is used as seed-parent. 



Thus, using B and M to stand for (E. biennis and (E. muricata, the 

 cross (B X M) x (M x B) gave plants resembling pure (E. biennis with 

 no trace of muricata : while (M x B) x(B x M) gave as offspring pure 

 muricata. 



The hypothesis adopted by de Vries to explain the above results is 

 that pollen grains and ovules carry different sets of characters. If the 

 characters carried by (E. biennis ovules and pollen grains are represented 

 hy B % and B ^, and those by (E. muricata hy M % and M ^, then 

 {B X M) has the constitution B% M ^, and is an entirely different 

 cross from {M x B) which has a constitution if $ B ^f. 



It is further supposed that there is selective sterility of half the 

 pollen grains and ovules in such a way that the ovules of the hybrid 

 {B X M) are of one constitution only, namely B $ — those of the con- 

 stitution M {/• being abortive. Similarly, the pollen grains are of one 

 kind only — namely those carrying M ^. 



The same scheme is applied mutatis mutandis to the reciprocal cross 

 {M X B). 



Thus it will be seen that on selfing {B x M) there will be a meeting 

 of male and female gametes carrying respectively M ^ and B % , giving 

 a zygote B % M ^. In other words the hybrid breeds true. 



1 •• Ueber doppeltreziproke Bastarde von (E. biennis und (E. muricata," Biol. Centralbl. 

 XXXI. p. 97, Feb. 1911. 



2 For other cases in which reciprocals differ in the direction of resembling male parents 

 see Focke. 



Papaver somniferum L. x P. caxicasicum M.B. as regards general size, colour of flowers, 

 &c. (Godron, Rev. d. Sc. natur. 1878, N. 2). 

 Probably flower colour in Petunia (Gartner). 



