Hymenophyllaceae, Schizaeaceae and Gleicheniaceae. 521 
by the quite similar endodermal pockets and the internal 
endodermis enclosing sclerotic tissue (although of local occur- 
rence), and as the form of the inner endodermis in N. dichotoma 
suggests reduction from a more extended tissue, the most 
satisfactory explanation of the facts is that the stelar structure 
of S. dichotoma is due to reduction from a type in which the 
endodermis had the same distribution as in Anemia mexicana. 
This view does not involve reduction from solenostely, but 
only from ectophlaic siphonostely, because there is no suf- 
ficient evidence that Schizaea or its ancestors ever had 
internal phloem. Nor does it directly touch the question of 
the morphology of the central tissues. That, on the present 
supposition, could only be attacked by an examination of 
forms (if such were extant) transitional in structure between 
a protostelic type and the supposed c siphonostelic ’ ancestor 
of the Schizaeas. 
The structure of N. dichotoma is, however, of interest in 
relation to the general question of the morphology of tissues. 
For, if one attempts to assign morphologic value to the endo- 
dermis as would be implied by using the word c phloeoterma,’ 
the natural inference is that the spindle-shaped sheath of 
inner endodermis, referred to above, contains cortical tissue, 
that in Fig. 25, E , where this endodermis is present, the bulk of 
the central tissue (that between inner endodermis and xylem) 
is stelar, and that, where the inner endodermis, after thinning 
off, comes to an end, all the central tissue must be regarded 
as stelar. It is difficult to see how an exponent of the 
morphological importance of the endodermis can avoid this 
conclusion. 
The writer’s view maybe stated here: (1) that the local 
complications of structure in the rhizome of S. dichotoma are 
reversional phenomena occurring at certain periods in the life 
of the plant, probably when nutritive conditions were at their 
best, as indicated by Thomas (’02, p. 344 et seq.), in the case 
of local reversional complication of sporophylls in Tmesipteris ; 
(2) that the structure of Schizaea has probably been derived 
from protostelic structure by the following stages, (a) appear- 
