114 Lillian O'Keeffe. 
Sablon (8) and figured by the latter writer. The lobing was in 
many cases very slight indeed, and cannot be described as similar 
to the deep lobing characteristic of the spore-mother-cells of the 
Jungermanniales. 
Many of the sterile cells (elaters) are seen to lie immediately 
within the single-layered capsule wall, in some cases forming a more 
or less definite layer in this region of the young capsule (Fig. 2, F). 
As stated by Kashyap (7), some of the elaters are in most cases 
found attached to both the upper and lower portions of the capsule 
(Fig. 2, H), though the number of these attached elaters is relatively 
small and their distribution (apart from the “ polarity” they show 
in occurring at the two opposite ends of the long axis of the capsule) 
irregular. In a few of the capsules sectioned very few, and occasionally 
none, of these fixed elaters could be found; the statement by Deutsch 
that none occur may perhaps be attributed to his having examined 
few capsules. In some places the sterile cells immediately within 
the capsule wall fail to separate from this at any point and conse¬ 
quently remain attached so that at such points the wall appears 
two-layered, but in most cases the separation is effected at an early 
stage, and when the developing elater fails to split completely from 
the wall-layer the attachment persists so that the mature capsule 
usually contains a number of these “ fixed elaters.” 
VI. Summary. 
1. The thallus of Targionia hypophylla grows by means of a 
single apical cell, from which segments are cut off dorsally, ventrally 
and on either side. The air-chambers arise by splitting between 
the epidermal cells, the split extending from the surface inwards 
through the entire depth of the epidermis ; this process is followed 
by partial closure of the crack due to turgor of the bounding cells ; 
the young chamber next undergoes extension owing to general 
growth of the thallus tissue, but remains closed for a considerable 
time until the divisions occur which cut out the concentric cell-rings 
around the pore, when the latter opens and continues gradually to 
increase in area until the general growth of the thallus ceases. 
The membrane-like rim around the pore of the chamber is formed 
from the innermost ring of surrounding cells, the cavity of these 
cells being almost obliterated by thickening of the walls. 
2. The antheridial receptacle may be developed on special 
short stalked disc-like branches of limited growth, or on ordinary 
thallus branches, every intermediate condition being found between 
