208 Evans: HeEpatTIcAE OF PUERTO Rico 
like that of the two Javan species of Radula, R. Hedingeri and R. 
Tjibodensis, described by Goebel.* In order to distinguish a pro- 
tonemal structure of this type from the similar protonemata some- 
times found in spore-germination, S¢ehiffner + proposes the term 
‘Brutknospenvorkeim’”’ or ‘‘gemmothallium,’’ which well ex- 
presses the facts in the case. In the gemmae of C. accedens the 
gemmothallia may apparently arise from any part of the margin. 
In the majority of cases, however, the apical region is favored, 
and it is possible under these circumstances that the old apical 
cell reassumes its function of dividing and becomes the apical cell 
of the gemmothallium. Next to the apical region one of the two 
basal cells seems to be the favorite starting-point for the new 
structure, and in some cases two or more gemmothallia may be 
found on a single gemma (FIGURE 23). From whatever region it 
arises the gemmothallium grows by means of a two-sided apical 
cell and finally equals or surpasses in size the original gemma. 
It affixes itself to the substratum by means of rhizoids like’ those 
found on the gemma, but owing to its flat character these rhizoids 
are not confined to the margin but can grow out from any part of 
the ventral surface. The apical cell of the gemmothallium event- 
ually becomes the apical cell of the leafy shoot. This gives rise 
at first to two very rudimentary leaves unaccompanied by an 
underleaf; the third and fourth leaves, however, begin to show 
the foliar characters of the species and are accompanied by an 
underleaf in the ordinary way. 
The gemmae of C. convexistipa are very similar to those of C. 
accedens and are attached to the leaf by two basal cells in the same 
way. They are, however, larger and commonly bear three or four 
unbranched marginal rhizoids instead of two. They are distin- 
guished further by a persistent apical cell which can be readily 
demonstrated in a mature gemma (FIGURE 14). Here again the 
gemma is normally inverted before it germinates ; it also affixes 
itself to the sustratum by means of rhizoids springing from its 
marginal cells. These rhizoids, however, do not arise from the 
ventral surface of the marginal cells as in C. accedens but from 
the outer edges, and are usually less branched (FIGURE 16). 
The gemmae of C. convexistipa do not develop gemmothallia but 
* Ann. Jard. Buitenzorg, 7: 51-54. p/. 6, 7. f 60-67. 1887. 
¢ Oesterr. Bot. Zeits. 43: 207. 1893. 
