508 Enans: HEPATICAE OF PuERTO RIco 
lobes are strongly squarrose and are sometimes completely 
reflexed, so that they come to lie parallel with the axis and with 
their apices pointing toward the base of the propagulum, The 
lobules are very rudimentary and in the reflexed leaves are 
turned forward, in this way obliterating the keel. The modified 
leaves, instead of being entire like ordinary leaves, are angular- 
dentate with from one to three teeth, one being at the apex, and 
the others (when present) on the outer edge. These teeth often 
end in the curious rhizoid-like processes (f r7) described by 
Schiffner * for Cheilolejeunea versifolia and for certain other 
hepaticae, but it is doubtful if they ever assume the function of 
rhizoids. As we pass from the base toward the apex of the pro- 
pagulum, we find all gradations between these modified leaves and 
the ordinary leaves of the species. ; 
The first two underleaves of the propagulum are perhaps even 
more remarkable than the leaves. They are larger than ordinary 
underleaves and their divisions often bend inward instead of 
diverging. The middle cells of the basal portion, which are more 
numerous than usual, do not produce rhizoids directly but develop 
instead, a large, orbicular, sucker-like disc (/ 72) composed of 
minute and closely united cells. This disc is entire on the marg!” 
and often projects slightly beyond the basal portion of the under- 
leaf. 
When a propagulum has developed several pairs of leaves, 1 
becomes separated from the parent plant and develops into 4 new 
individual. The separation takes place at its junction with te 
parent-axis just within the basal sheath. The cells by which it 1S 
actually attached are few in number, often reduced apparently to 
a single one. These cells gradually become shriveled, and it re- 
quires very little force to tear them across and set the propagulum 
free. Probably in nature the force of a shower would be sufficient 
to bring about this result. It has already been stated that a pro- 
pagulum, while still attached, lies withits underleaves turned away 
from the substratum. This position cannot be very stable after 
separation takes place, because the propagulum would rest (a 
nothing except the edges and apices of its squarrose leaves. Pro : 
ably here again the force of dripping water is sufficient to Ove 
* Engler’s Bot. Jahrb. 23 : 598. p/. 25. fg. 1897. 
