XX] 



GAMETOPHYTE 



119 



prothallus does not start with a filamentous stage, but at once proceeds to 

 a flattened expansion, soon forming the massive cushion. In old prothalli 

 apical dichotomy may appear, as in many Liverworts {E, F). Campbell 

 recognises a near resemblance to the prothalli of Osmiinda, and especially 

 in the tendency of both for the prothallus to be more than one cell thick 

 from the first. In the Marattiaceae this becomes very conspicuous with age, 

 extending almost to the margin of the thallus. Further, adventitious buds 

 originate from the margin, often forming independent secondary prothalli. 

 The rhizoids have been described as being septate: but Campbell finds in 

 CJii'istensenia that this does not result from a true cell-division, since only 

 one nucleus is found in the whole rhizoid. 



In all the genera Campbell found an endophytic fungus occupying the 

 central regions of the prothalli, as in the Ophioglossaceae. The mycelium 

 resembles that of the latter, but in the Marattiaceae there is no evidence of 

 digestive cells containing the characteristic varicose mycelium. It appears 

 that here there is no destruction of the fungus by the host: it seems to live 

 more nearly as a true parasite than in the saprophytic Ophioglossaceae. 



a 



Fig. 411. a—d= Marattia Douglasii, after Ca m pbell. e—f= Ophio- 

 glossiim pendulum, after Lang. a = antheridium, with divisions of 

 spermatocytes (32) in section, perhaps not complete; ^ = young 

 antheridium ; f = archegonium; d—yawng archegonium; t-^anthe- 

 "■^■' ''^ °° spermatocytes in section; /= archegonium. 



ridium with 



