Bower.—Studies in the Phytogeny of the Filicales. VII. 33 
• 
some fluctuation in detail. The segmentation of the sterile pinnule is quite 
typical. But where a sorus is to be formed the curved outline of the 
young margin becomes slightly irregular, as seen in the transverse section 
(Fig. 24, A, b), and the marginal cell appears to be diverted downwards 
(i. e. to the abaxial surface), owing to stronger growth of the last adaxial 
segments. Its products soon grow into a projecting ridge, in section conical, 
which appears then to lie upon the lower surface (Fig. 24, c). The divisions 
in the last adaxial segments may simulate the marginal segmentation 
itself, so that it becomes difficult to trace the identity of the real margin. 
It is possible that in some cases there may have been a slide of the ridge 
actually to the lower surface. But comparison of a large number of 
examples leads to the conclusion that the ridge (or in section the cone), 
which thus appears to be on the lower surface, is essentially, even if not 
always actually, of marginal origin. It develops as the receptacle (r in 
Fig. 24, D, e), while the stronger growth of the adaxial segments has 
produced the upper indusium (u y Fig. 24, D, e) ; the latter is of similar 
origin to that in Pteridium or Paesia , or the Dicksonioids generally. As 
growth proceeds, these parts become more definite in form (Fig. 24, f), 
and the apparently superficial position of the receptacle more pronounced. 
From the actual apex of the ridge certain cells project ( s , Fig. 24, G, H). 
These are the first sporangia, and if the above account of the origin of the 
receptacle is correct, then these first sporangia are, as in so many of the 
Marginales, derived from the marginal cells themselves. There is, however, 
no inner indusium. This was a matter of old observation in the mature 
state, and was in fact the justification of the inclusion of the species in 
Pteris , \Histiopteris , now held as a substantive genus. Later sporangia 
show indications of a basipetal sequence, which, however, is not strictly 
maintained (Fig. 25, i, ii). With the sporangia hairs (paraphyses) are 
scattered, and sometimes one may accidentally lie between the receptacle 
and the lower surface, giving the appearance of a lower indusium (Fig. 
25, ii); but this is quite inconstant, and more commonly none is seen 
there (Fig. 25, i). A vascular commissure runs below the receptacle; it is 
almost circular in section, as in Pteridium and Paesia y and the receptacle is 
clearly conical in form. In these characters it maintains its similarity to 
the type of the less advanced Marginales. 
These observations indicate what has probably been the phyletic 
history, that Pteris ( Histiopteris) incisa has probably sprung from some 
Lindsayoid-Paesioid source with double indusium. The steps in advance 
were reticulation of the veins, formation of dermal scales, loss of inner 
indusium, and a tendency to slide the sorus on to the lower surface ; 
this last comes out more clearly in the mature than in the young state. 
On the other hand, the anatomy is simpler than in Pteridium , but 
more complex than in Paesia. This has been already noted by 
D 
