166 EMBRYOGENESIS IN PLANTS 



another and lead to its precocious development, or to its delayed or 

 inhibited development. This, indeed, is what seems to happen in fern 

 and lycopod species with saprophytic prothalli. 



In the normal development, as in the embryogeny of Dryopteris, 

 Onoclea or Adiantwu, the primary organs, shoot, leaf and root, are 

 formed almost simultaneously and grow in a mutually regulated or 

 harmonious manner; and in all the subsequent growth and organo- 

 genesis, including the sustained growth of the adult over many years, 

 this developmental harmony is apparent, i.e. a specific pattern of 

 development is maintained throughout. The embryonic developments, 

 in fact, lead on naturally and conformably to those that follow. But, 

 in that the adult fern plant is a considerably more complex structure 

 than the embryo, it may be inferred that in the ontogenetic development 

 there is a continuous elaboration of the underlying biochemical pattern. 



The Embryo a Spindle or Filament. In all fern embryos polarity is 

 determined at the outset. In species with a suspensor, the young 

 embryo typically undergoes some elongation and is filamentous in 

 character. Where no suspensor is present, the zygote becomes an 

 enlarging sphere, or spheroidal body, dividing successively into hemi- 

 spheres, quadrants and octants. In species with a suspensor, the distal 

 embryonic cell develops in an essentially similar manner. From a 

 general survey of plant embryos, especially those of Bryophyta and 

 Pteridophyta, Bower concluded that all were initially more or less 

 spindle-like and were 'based on the uUimate type of a transversely 

 septate filament' (1923). In his view, the two-celled stage of the lepto- 

 sporangiate fern embryo could be regarded as a filamentous stage, this 

 reflecting a primitive ancestral state. While there may be substance in 

 such a view, we should, perhaps, be chary of accepting it out of hand. 

 The shape of a growing zygote depends on allometric growth, i.e. on 

 the relative growth development in diff'erent directions. If it grows 

 rapidly along one axis only, a spindle, or filament, will necessarily 

 resuU, and this will normally tend to divide by transverse walls. If its 

 rate of growth is equal along all three axes it will become spherical; 

 and equatorial, quadrant and octant divisions are likely to follow. 

 Until more is known of the nature of the reaction systems that yield the 

 different distributions of growth, views as to the relative evolutionary 

 status of filamentous and spherical embryos must necessarily remain 

 speculative. Nevertheless, it is a fact that the most filamentous and 

 alga-like fern embryos, i.e. those with a suspensor, belong to euspor- 

 angiate species which on other grounds are held to be of ancient and 

 primitive stock. As Bower has pointed out, no leptosporangiate fern 

 is known to have a suspensor and there is no evidence of the formation 

 of that organ de novo. The Ophioglossaceae and Marattiaceae aff"ord 



