112 



MOSSES AND FERNS 



CHAP. 



number of Liverworts belonging to all the principal groups 

 have been examined by me, and invariably the presence of a 

 definite cell wall could be demonstrated at all stages. 



Many of the foliose Hepaticge show much greater regu- 

 larity in the early divisions of the embryo, and in the establish- 

 ment of the archesporium and the arrangement of its cells. 

 This is especially marked in Frullania (Leitgeb (7), II.). 



Here, after the upper part of 

 the embryo has divided into 

 three tiers of cells, these under- 

 go the usual quadrant divi- 

 sions, and the four terminal 

 cells only, form the capsule, in 

 which the archesporium is es- 

 tablished by the first periclinal 

 walls (Fig. 58). The divi- 

 sions in the archesporium are 

 also extremely regular, so that 

 the spores and elaters form 

 regularly alternating vertical 

 rows. In Frullania the lower 

 cell of the embryo, instead of 

 remaining undivided, or form- 

 ing simply a row of cells, di- 

 vides repeatedly, and the cells 

 grow out into papillae, so that 

 it probably is functional as an 

 absorbent organ, like the foot 

 of the Anthocerotes. Radula 

 (Hofmeister (i)) and Junger- 

 mannia, while more regular in 

 „,,„,,. ^ . the divisions than Porella, still 



Fig. 57. — Porella Bolandert. Longi- . 



tudinai section of a sporogonium after are Icss SO than Frullama, and 



the f^nal division of the archesporial j^ thcSC mOrC than the Upper 



tier of cells take part in the 

 growth of the capsule. The degree to which the seta and 

 foot are developed varies. In Porella there is not a distinctly 

 marked foot, the lower part of the seta being simply somewhat 

 enlarged, but in others, like Jungennannia hicuspidata, there 

 is a large heart-shaped foot, very distinct from the seta. In 

 Porella the seta is short, projecting but little beyond the 



