EMBRYOGENESIS IN FLOWERING PLANTS 287 



material to be obtained. Haploid parthenogenesis can be induced in 

 some species by the use of appropriate techniques {see Chapter XV), 

 e.g. by stimulating but not fertilising the ovum with pollen of another 

 species, by delayed pollination, etc. Jorgensen (1928) found that when 

 the stigma of Solanum nigrum was pollinated with the pollen of 

 S. hiteum, the male nucleus eventually penetrated the ovum, but there 

 was no effective nuclear fusion and the male nucleus soon began to 

 disintegrate. The ovum, however, was activated as if by normal 

 fertilisation and an embryo was formed. Twin proembryos may some- 

 times be observed in Liliiun martagon, Fig. 79a ; one of these is formed 

 from the ovum by the normal fertilisation, the other from a stimulated 

 haploid synergid cell. The initial developments of the two embryos are 

 closely comparable but the synergid embryo soon degenerates (Cooper, 

 1943). Similar observations on synergid embryos have been made by 

 Lebegue (1949) in Bergenia delavayi (Saxifragaceae) and by Crete (1949) 

 in Erythraea centaurium (Gentianaceae). Maheshwari (1950) has sum- 

 marised other evidence on parthogenesis, including cases of androgenic 

 haploids in which, as the evidence seems to show, the embryo has 

 developed from the male nucleus alone. In other instances of apomixis 

 the chief interest lies in the cytological rather than in the embryonic 

 vicissitudes. Here we may note that although there may be departures 

 from the normal nuclear arrangements, the embryonic development, 

 presumably in relation to the relatively stable conditions within the 

 ovule, initially follows normal paths. Where there is subsequent 

 degeneration, this is usually attributable to factors in the genetical 

 constitution. 



In Rudbeckia speciosa (Compositae) a peculiar embryonic innova- 

 tion, described as semigamy, has been observed. Fig. 79b-d (Battaglia, 

 1946, 1947). The small male nucleus enters the ovum which is diploid, 

 and activates it but there is no nuclear fusion. On the contrary, it 

 appears that the male and female nuclei divide separately, though often 

 simultaneously during the ensuing embryonic development, with the 

 result that cells containing male nuclei, which are never numerous, 

 come to occupy variable positions in the embryo, a sort of embryonic 

 chimaera being thus formed. 



In adventive embryonic development, the embryo has its inception 

 in a diploid cell of the nucellus or integument. This cell becomes densely 

 protoplasmic and divides actively to form a small mass of meristematic 

 cells. This tissue mass grows, or is pushed into the embryo sac, and 

 there it gives rise to what appears to be a normal embryo. In Citrus 

 trifoliata (Rutales: Rutaceae), Fig. 79e, f, the normal embryo has a 

 suspensor whereas the nucellar adventive embryos have not, i.e. 

 they resemble the distal region of the normal embryo. From these 



