104 MISSOURI BOTANICAL GARDEN. 
such families coming in contact would readily interlock 
and thus form the beginning of a mat of the plant at the 
surface. This mat would be intensified as the young fronds 
push their curved tips downward through the mass and 
lock the families in closer embrace. This will account for 
the dense masses found in many places.* 
The individual plant itself is an elongated strap-like 
frond, about fifteen times as long as the broadest part and 
but a few cells in thickness, tapering from the oblique base 
with its broad pouch-cleft to a narrow rounded apex which 
is only about } to }.as broad. The margin is entire. As 
stated in the introduction, the flowers and fruit of this 
plant are not known, but it reproduces quite freely by a 
form of branching or budding. At the basal end of each 
frond the two epidermal layers are split apart to form a 
deep elongated triangular pouch which extends about } to 
¢ the length of the frond (pl. 64,f. 7). Atthe bottom of 
this pocket, in the acute angle, the budding takes place 
which results in new individuals being cast off to take up 
a life’s work in themselves. The young frond grows out 
from the pouch-opening in the same general direction as 
its parent, until of about the same length, when it takes a 
position at about right angles to its parent, by which time 
it also is sending out a young frond. In fact a frond is 
scarcely more than half matured before it begins to form 
buds in its pouch, while at the same time its parent is send- 
ing out other individuals — sister fronds — alongside it. 
Each frond is attached to its parent by an elongated stipe 
which is long persistent, thus holding individuals together 
for many generations, forming the families above referred 
to (pl. 64, f. 6). 
In structure the frond consists of the two epidermal 
layers separated in the greater part of its length by large 
air chambers and united only at the margins and a small 
portion of its apex. They are connected in the cavernous 
* John Donnell Smith: Bull. Torr. Bot. Club, vir. 64 (1880). 
