192 THE ALGAE 



into a discoid holdfast, stipe and frond, and in all genera except 

 Durvillea, growth in length is due to the activity of one or more 

 apical cells. The genus Bifur carta is anomalous in that the discoid 

 holdfast is replaced by a basal rhizome. Morphologically an exter- 

 nal meristoderm, cortex and inner medulla can be recognized. In 

 the genera commonly occurring in the inter-tidal regions the walls 

 of the central or medullary cells become abundantly gelatinized 

 whilst there is a greater profusion of longitudinally running hyphae, 

 though Hormosira is an apparent exception to this in that the 

 swollen segments become hollow. In the genus Halidrys, sieve 

 plates occur on the walls of the medullary cells and there is an in- 

 dication of similar plates in the lateral medullary connections of 

 Bifurcaria. Reproduction is by means of oogonia and antheridia 

 borne in conceptacles which, in all but Durvillea, are locaUzed in 

 regions known as receptacles. Some workers consider that the 

 structures called oogonia and antheridia are really macro- and 

 microsporangia producing mega- and microspores which germinate 

 before they are liberated from the sporangium, so that while the 

 reproductive bodies have their origin as spores, nevertheless the 

 liberated products are gametes. This view is discussed more fully 

 later (cf. p. 317). In the primitive condition eight ova are produced 

 in each oogonium and sixty-four antherozoids in each antheridium. 

 Meiosis takes place during the first two divisions in the formation 

 of microspores, and as there is often a pause after the second divi- 

 sion the first four nuclei have been regarded as the functional micro- 

 spores, each of which subsequently undergoes four mitoses so that 

 they can be said to germinate to a sixteen-celled gametophyte where 

 each cell functions as an antherozoid. In the macrosporangium the 

 first four nuclei formed have been regarded as the functional mega- 

 spores, and each of these is considered to germinate subsequently 

 to a two-celled female gametophyte where each cell functions as an 

 ovum. In those species where less than eight mature ova are pro- 

 duced it must be assumed that some of the megaspores undergo 

 abortion or else do not develop. 



If the above is to be the correct interpretation, and it would seem 

 to be more satisfactory than any other theory in comparison with 

 other members of the Phaeophyceae, then we can say that not only 

 is there a cytological alternation of generations but there is also a 

 morphological alternation, although the sexual generation is even 

 further reduced from the state found in the Laminariales. This 



