460 Bower. — Studies in the Phytogeny of the Filicales. 
appear, and the largest constantly occupies a position which is at least near 
to, if not actually coincident with, the margin of the pinna. Other sporangia 
appear later in lateral positions, mostly on the adaxial side. Hairs are also 
formed in considerable numbers. A transverse section of a sorus at this 
stage is shown in Fig. 19, and from such a section it is clearly proved that 
the sorus, though continuous and marginal, is essentially a gradate one in 
its early stages. The oldest sporangia form a fairly regular series in 
a central position between the indusial lips, while those which follow are 
laterally placed. It will be noted that the early segmentation of the 
sporangia is such as to produce a fairly massive stalk, and resembles that 
of the Gradatae generally. A further point for note is the irregular lobing 
of the indusial lips (Fig. 18). The laciniae thus produced are chiefly upon 
the lower lip, and they are a more marked feature in O. retusa than in 
L . lancea. It is also seen that while the lips are only slightly unequal in 
O. retusa , and the sorus consequently appears terminal, they are much more 
unequal in L . lancea (Fig. 20), and the sorus is distinctly intramarginal 
(Fig. 21). The sporangia are closely packed between the lips in the latter 
Fern, and arise in the later stages of the sorus after the manner of the 
Mixtae. But the sequence is not long continued ; in fact, neither Odonto- 
soria nor Lindsay a appear to be at all pronounced cases of the mixed con- 
dition of the sorus. 
It is thus seen that the sori of the Ferns above described are in an 
intermediate condition between the Gradate and the Mixed types. The 
receptacle, usually a projecting body in the Gradatae, has become flattened 
(Fig. 18), and the basipetal sequence, though it can be traced in the younger 
stages, is not strictly adhered to. This at once suggests the question 
whether the position of the annulus is oblique, as in the Gradatae, or vertical, 
as in the Mixtae. A similar question would apply to Saccoloma. Unfortu- 
nately, the Jamaican material did not suffice to decide the question for that 
Fern or for Lindsay a lancea , though in both cases evidence of an oblique 
position of the annulus was seen. But examination of the sporangia, both 
in Odontosoria retusa , (Cav.) J. Sm., and in Lindsaya repens , (Bory) Bedd., 1 
showed that in certain sporangia the annulus was clearly oblique, and the 
sequence of its cells was not interrupted at the insertion of the stalk. There 
is, however, some variety of detail in different sporangia. In Fig. 22 an 
example is shown of the sporangium of Lindsaya repens , in which, though 
the induration of the ring stops opposite the insertion of the stalk, still the 
sequence of cells is continued past it, so that a complete oblique ring is 
present, as in the Dicksonieae. A similar condition has been observed in 
Odo 7 itosoria retusa. Here again, though the induration ceases opposite to 
the stalk, the series of cells of the annulus is continued past it. This con- 
1 This plant is named in the Synopsis Filicum Davallia (Odont.') repens , Desv., and Hooker 
there notes that its place is ‘quite doubtful between Odontosoria and Lindsaya \ 
