Bower.—Studies in the Phytogeny of the Fiticates. 439 
the structure may, however, be simpler in the lower region of the stalk. In 
the sporangial head the segmentation to form the tapetum follows the 
usual plan (Fig. 16,/, g), and it later divides into two layers (Fig. 16, d), 
and becomes disorganized in the usual way (Fig. 17, c , d). Meanwhile, the 
cell remaining at the centre divides to form the spore-mother-cells, apparently 
twelve in number, and the further development proceeds in the usual way. 
The total output of spores per sporangium appears to be forty-eight. This 
is a feature which Plagiogyria shares with many of the Gradatae and 
Mixtae, but it is to be noted that a limited spore-number (48-64) is also 
seen in Matonia , while in the Schizaeaceae the spore-numbers are not high 
as compared with what is seen in others of the Simplices. The spores 
themselves are tetrahedral, and when fully ripe numerous highly refractive 
globules are found attached to their external walls; the attachment is not 
a very close one, as is shown by the fact that they are easily separated in 
the mounting, and may be found lying in the mounting medium apart from 
the spores; they are formed just before the maturity of the spore. Apart 
from the globules the wall of the spore has no very distinctive markings ; 
in general character it shows some similarity to that of Lygodium (Engler 
and Prantl, i, 4, Fig. 192, j), but the spore-characters cannot be held to be 
altogether trustworthy for comparison among the Simplices, as both in 
Schizaeaceae and Gleicheniaceae they are variable within the order. The 
tetrahedral type is, however, the prevalent one among the early Filicales. 
The mature sporangium has been depicted by Mettenius ( 1 . c., Figs. 2 
and 3), and his drawings have been copied by various later authors ; they do 
not, however, give all the detail for necessary comparison with those of other 
Ferns, especially when it is remembered how important a part the sporan¬ 
gial structure plays in arriving at conclusions as to phyletic relationships. 
The mature sporangium is long-stalked (Fig. 18, a, b), in fact quite 
unusually so among Ferns with oblique annulus, in which the sporangia are 
as a rule short-stalked, or sessile. A comparison in this respect may be 
made, on the one hand, with the Osmundaceae, and ultimately with some 
Botryopterideae ; on the other, with Gleichenia § Mertensia , and especially 
with G. dichotoma (‘ Land Flora/ Fig. 310, and text, p. 556); with the latter 
there is also correspondence in the absence of any central cell or cells in the 
transverse section of the stalk, which is an indication of the less massive 
structure of the sporangium. There is a gradual transition from the spo¬ 
rangial stalk to the head, giving the whole body that pear-like form which is 
so conspicuous in G. dichotoma ; it appears also in Loxsoma (‘ Land Flora/ 
Fig. 320, p. 572), and in certain species of Dicksonia (‘ Land Flora,’ Fig. 330, 
p. 593). The head is traversed obliquely by the annulus, which is a complete 
ring, as recognized by the sequence of the cells, though there is variety in the 
manner of their induration ; there is also some irregularity in their form and 
number—thirty being about the average; there is also some variability in 
