290 



PROTOCYATHEACEAE 



[CH. 



but that o{ Mctaxya appears as a wide gutter. It has been seen that the stele 

 of LopJwsoria has advanced towards dictyostely in the shortened upright 

 stock, with a tentative disintegration of the meristele near to the leaf-base. 

 But notwithstanding these differences of detail in essentials the system is 

 comparable in all of these ferns. They are all advanced from protostely to 

 solenostely, but have not arrived definitely at dictyostely or disintegration. 

 The sori of Metaxya are oval in outline, with a flat receptacle, upon which 

 50 to 100 sporangia are closely grouped, together with long occasionally 

 branched hairs. The number of the sporangia is in strong contrast to the 

 16, or often much less, in each sorus of G. pectinata, LopJwsoria, and the 

 other Gleichenias. The development of them all is simultaneous, as in the 

 SimpHces, to which, including Metaxya, they all technically belong. The 

 orientation of the sporangia in Metaxya is not uniform: they have an almost 

 vertical annulus of about 27 cells: it is as a rule interrupted at the insertion 



four aspects of the sporangium. 

 For details see text. ( x 100.) 



of the stalk, and it is almost vertical with a well organised stomium (Fig. 

 553). The slit of dehiscence is in a transverse plane, which is convenient in 

 a flat sorus where the sporangia originate simultaneously. In this it differs 

 from the Gradate Cyatheaceae, where the slit is oblique. The spore-output 

 is 64. An interesting feature is the four-rowed stalk, giving a characteristic 

 transverse section (Fig. 554). This may be held as an indication of the mode 

 of segmentation of the sporangium. It would naturally follow from a two- 

 rowed segmentation, in place of the three-rowed which is usual in advanced 

 Ferns. This proves to be the fact in Metaxya, a point of some comparative 

 importance (Fig. 555). Nevertheless the annulus is formed just as in 

 sporangia with the three-rowed segmentation, and it provides a pregnant 

 instance of the independence of segmentation from morphological initiation 

 (see Vol. I, p. 246). The gametophyte of Metaxya is unfortunately unknown 

 These features provide interesting material for comparison with a view 

 to phyletic seriation. In respect of form Metaxya retains the creeping 

 habit of the Gleichenias, and has not like Lophosoria assumed an 



