196 BOTANICAL GAZETTE [MARCH 
matic region was first visible in the seventh leaf (figs. 23, 31, 34). 
Such a region has been reported by GWYNNE-VAUGHAN (g) to occur 
in Kaulfussia andArchangiopteris, and WEst (13) reports the same for 
two species of Danaea. Although not reported for the other genera 
of the Marattiaceae, it is probable that this region is found in them 
also. GWYNNE-VAUGHAN suggests that this region might represent 
the rudiments of the adventitious buds, and WEstT reports that this 
does actually occur in Danaea nodosa. While this region may give 
rise to adventitious shoots, it seems probable that its principal 
function is to build up the fleshy stipules. In none of the specimens 
of Angiopteris studied was there any evidence of the presence of 
adventitious buds in this region. 
Conclusions 
It is evident that Angiopteris presents an example of striking 
general variation in the vascular structures from stage to stage. 
How far this variation is continued remains to be determined from 
further studies of more advanced stages than have been dealt with 
in this investigation. From the facts observed in the stages 
studied, however, it seems probable that some of these variations 
are continued indefinitely as the plant increases in size and age. 
The most conspicuous of these variations are: (1) the elimination 
of the endodermis, (2) the appearance of commissural and medullary 
strands and the increase in their importance in the structure of the 
central region, (3) the repeated bifurcation of the leaf traces, (4) 
the placing closer to the central strand the point of attachment of 
the commissural strand and the point of forking of the leaf trace, 
and (5) the variation in the place of attachment of the root steles. 
Much of this variation tends to the breaking up of the central 
strand, a fact which points to a polystelic condition so characteristic 
of the phylogenetically advanced types of the different groups of 
vascular plants. 
The total absence of a cauline procambium is interesting, as it 
suggests that the centra! vascular structure in the stem of the 
Marattiaceae, and perhaps other closely related ferns, is mostly 
if not wholly of foliar origin; that is, it consists of a sympodium of 
leaf traces. The definite relation between roots and leaves strongly 
supports this theory. 
® 
