EMBRYOGENESIS IN FLOWERING PLANTS 255 



embryos observed being small and undifferentiated, or at the incipient 

 cotyledonary stage. The full details of the embryonic development have 

 yet to be ascertained. A somewhat similar account has been given of 

 the seed of the vesselless dicotyledon Amborella trichopoda (Bailey and 

 Swamy, 1948). Amborella was originally placed in the Monimiaceae, 

 but Bailey and Swamy consider that this requires much more evidence 

 before it can be accepted. The embryogenies of the nine surviving 

 genera of vesselless dicotyledons merit investigation, {see Bailey and 

 Nast (1945) for a discussion of the Winteraceae, etc.) 



In Cercidiphyllutn japonicum the zygote divides transversely and the 

 terminal cell again divides transversely and then longitudinally. The 

 basal cell becomes a swollen suspensor cell, while the terminal cell 

 gives rise to a globular cell mass. Fig. 62j, l (after Swamy and Bailey, 

 1949). Meanwhile the embryo sac has elongated greatly. The cellular 

 endosperm is characterised initially by dense protoplasm at the chalazal 

 end. This genus was formerly associated with several others {Trocho- 

 dendron, Tetracentron, etc.), but the anatomical evidence, including the 

 embryology, 'provides no cogent arguments for including this genus in 

 any particular family of the dicotyledons.' It should be placed in an 

 independent family on its own. There are no very close resemblances 

 between the embryos of Cercidiphyllum and of Degeneria. (For a 

 discussion of the taxonomic relationships of primitive dicotyledons, 

 see Swamy and Bailey, 1949.) 



Sassafras variifoliiim (Laurales), a primitive dicotyledon, belongs to 

 a genus with as long a fossil history as any angiosperm known in 

 N. America (Coy, 1928). In chronological records it is preceded in time 

 by Populus, Liriodendron and Magnolia. The embryo sac is of the 

 normal eight-nucleate kind and after fertilisation there is rapid endo- 

 sperm formation. The zygote may divide either by a transverse or by 

 a longitudinal wall, i.e. the plane of division is not definitely fixed. In 

 its subsequent development the proembryo tends to be massive and 

 flattish, with a somewhat inconspicuous short wide suspensor. Fig. 

 59k, m. The mature embryo has massive cotyledons with auricles, a 

 plumule bearing small leaf primordia, and a flattish radicle. The 

 development is thus on a massive basis. Johansen (1950) states that 

 this embryo shows close agreement with the Scabiosa variation of the 

 Piper ad type. In Leitueria fioridaua the first wall of the zygote may be 

 transverse or vertical. The embryo develops a massive suspensor and 

 the pear-shaped embryo consists of hundreds of cells before there is 

 any evidence of cotyledon formation (Pfeiffer, 1912). This genus is 

 difficult to associate with any specific group in the Archichlamydeae. 



In Moringa oleifera (Capparidales) an embryonic development not 

 unlike that in Degeneria has been observed (Puri, 1941). There is 



