HEVUE BKYOLOGIQUE 83 



n 



of the axis (stem) shows a spiral arrangement of tlie cells^ siig- 

 gcstivc of lorsion, and tliat the axis ends below in a bulbous dila- 



w 



tation. The writer has not observed either of Ihese features in 

 /?. capensis. Here the colls of tlie stem tend to be arranged in 

 fairly regular and straight longitudinal rows and the lower extre- 

 mity is pointed. The wing consists for the greater part of a singK) 

 layer of cells^ but the portion lying nearest the axis is usually 

 two-layered, and of course in the male plant there is a well 

 marked tliickening at the margin of the wing, in which the anthe- 

 ridia are embedded. The oil-bodies present in many of the colls 

 of the wing, especially near the margin resemble those of the 

 Marchantiacetje, a single oil-body being found in a cell, which it 

 almost fills. 



The leaves are arranged in a fairly regular alternate manner 

 on each side of the axis. Each leaf is broad and semicircular, 

 cordate, or reniform in outline, and at about the middle of its 

 free margin there is a clear projecting papillate cell, reprcscnling 

 the « primary papilla » with wdnch llie young leaf begins its 

 growth. In R. Battandieri,Sc\\\i\nQY (i) has described and figured 

 the occurrence of leaf-like appendages on the marginal portion of 

 the wing, and in the allied It. amcrkana, discoid gemma) are 

 borne on the axis (2). In IL capensis^ neither of these structures 

 have been observed. 



Each branch of a male plant bears a very large number of 

 antheridia in a single, or here and there double, row along the 

 thickened margin of the wing. As many as 100 antheridia or 

 more were seen on some branches in an uninterrupted scries, 

 but frequently after a certain number have been formed ther(^ is 

 a break, and a thin sterile area intervenes before Hhe formation 

 of antheridia is resumed. In such cases, the posterior series 

 shows only the empty chambers, the antherozoids having been 

 discharged. As a rule, each anlheridium is sunk in a separate 

 cavity, but sometimes two antheridia lie side by side in the same 

 cavity. Each antheridial cavity is prolonged above into a narrow 

 canal, whieli runs obliquely forwards and opens by a pore on 

 the edge of the wing, the cells immediately round the pore pro- 

 ject as papilla?, and each pore occupies a slight conical promi- 

 nence, so that the margin of the wing is serrate as seen in profile. 



(1) SchifTner, Observationes de exoticis quibupdam Ilepaticis. Bot. Ccn- 

 tralb., 1886; Taf. 1. Fig. 5 ; Rev. bryoL, 1887, p. 113. 



(2) Howe and Underwood. The genus Riella, with descriptions of new 

 species from North America and the Canary Islands. Bulletin of the Torrey 

 Bot Club, 1903, p. 210, 



