1915] PETRY—OPHIOGLOSSACEAE 357 
in quantity at this location; hence this tissue is not an intrusion of 
the cortex, but is stelar in character. It is to be noted that this 
tissue is opposite a break in the endodermis, but not outside it. 
The view is held that there is a morphological distinction 
between stelar and cortical tissues; that the suberization of the 
radial walls of a layer of cells is a physiological phenomenon which 
under the usual conditions of development occurs in the layer of 
cells next outside the stele; and that this suberization, under the 
usual conditions, may be considered an indication of the morpho- 
logical boundary of the stele, but that it is subject to variation with 
physiological conditions. LAnc has concluded that the internal 
endodermis of B. Lunaria is of physiological significance only, and 
has suggested that its development is associated with the long leaf 
gaps of the intermediate region of the rhizome. It is to be noted 
that the apparent internal endodermis in injured specimens of B. 
obliquum (fig. 6, E, F) occurs between masses of vascular elements 
and points of injury. 
Of the stelar tissues, the tracheids and sieve tubes are incapable 
of further growth. The remaining tissues are separated by these 
into three groups: (1) pith, (2) cambium and adjacent parenchyma, 
and (3) pericycle. The data given above show that all these three 
tissues may produce vascular elements, either tracheids or sieve 
tubes. Hence we may conclude that the production of vascular 
elements by any stelar tissue is limited only by the capacity of the 
tissue for further growth. 
The manner of that further growth of a stelar tissue varies with 
the species. In Ophioglossum pendulum, which has no secondary 
thickening, the strands of xylem in the pith of an aberrant speci- 
men were primary in origin. In Helminthostachys, no secondary 
xylem is formed under the usual conditions of development; and 
renewed growth of the stelar tissues, as in the case of branching, 
produces an irregular secondary thickening of the stele without a 
definite meristematic layer. On the other hand, B. virginianum and 
B. obliquum show very great development of secondary wood under 
usual conditions. In both these species, renewed growth of the 
pith, cambium, and pericycle manifested itself in part by the 
formation of secondary xylem. Hence we may conclude that the 
