i 9 i6] STARR— AY TON I, I 53 



remaining attached for a longer time. The split may extend almost 

 immediately to the surface (fig. 15) or be delayed until the forma- 

 tion of the pore (fig. 16). The structure of the pore is more elab- 

 orate than in cavities of the thallus, as noted by Campbell (3) in 

 archegonial receptacles in general. The cells destined to form the 

 margin become evident by an increase in size (fig. 15); segments 

 are cut oil from the inner side, generally by oblique walls (fig. 17); 

 more divisions occur until 3-5 rows of cells form a ring extending 

 down into the enlarging cavity (fig. 18); then segments are cut 

 off from the outer part of the cells of the margin, walls being again 

 oblique, followed by another division, so that a projecting ring of 

 cells three layers thick is formed (fig. 19). The cells of the last 

 row shrink and become highly cutinized, giving a beaklike effect 

 in section (fig. 20), as noted by Deutsch (8) in Targionia. I find 

 no such thickening of the walls as Leitgeb shows in surface view, 

 but only such an effect as appears in fig. 21, taken from the older 

 part of the thallus. Of course my material may have grown under 

 conditions entirely different from his. With the increase in size 

 of the cavity the pore spreads and becomes a barrel-like opening. 

 The development is not complete until fertilization has taken place 

 and the receptacle has increased greatly in size. 



Sex organs 



The development of the antheridium is probably such as is 

 general in the Marchantiaceae, but my material has no early stages. 

 Occasionally the antheridium has a long stalk; less frequently the 

 upper end is beaked ; but the usual form is short-stalked and coni- 

 cal. About the antheridium are glandular hairs, and among the 

 cells forming the wall of the pit are mucilage cells, the secretions 

 from which probably assist the dispersal of the sperms. The 

 antheridia are produced in acropetal succession, so that the younger 

 are toward the apex (fig. 12). 



The archegonia are reported by Campbell (5) to be 3 or 4 

 in number, the one to the rear developing first, the lateral ones 

 next, and the one in front last. My material confirms the state- 

 ment in general, but 5 archegonia may occur. They begin develop- 

 ment near the top of the receptacle (fig. 14), but are carried under 



