OSMUNDACEAE 535 
is believed to be truly comparable to the “central cell” of the Marattiaceous 
‘sporangium. 
The archesporial cell thus surrounded by the tissue which will form 
the sporangial wall undergoes segmentation to form the tapetum (Fig. 296 © 
and F), and the irregularity seen in other segmentations is fully maintained 
in this also. From a comparison of Figs. 296F and G it will be seen that 
sometimes the segmentation of the tapetum is almost according to the 
ordinary Leptosporangiate type (Fig. 296G); the most interesting cases are, 
however, those in which the archesporium is truncate at the base; in 
these it appears that the cell or cells below it contribute to the completion 
of the tapetum (Fig. 296 F); this is a point of some special interest, for one 
of the most constant differences between the Leptosporangiate and 
Eusporangiate Ferns has been in the origin of the tapetum: but TZodea 
occasionally shows an intermediate condition. The division of the tapetum 
into two or partially three layers follows 
(Figs. 296 F to 1), together with the sub- 
division to form the sporogenous group ; 
beyond this point it will be unnecessary 
to trace the development of Zodea barbara. 
In conclusion the drawings not lettered in 
the centre of the group are added, as 
showing differences of thickness and seg- 
mentation of the sporangial stalk; these 
are both from sporangia, of the age of 
Fig. 296 H. 
An abnormality, which has been ob- Bigveaee Gs 
4 1 Osmunda regalis. A synangium, in 
served occasionally in Zodea barbara, has el Up a 
been found not unfrequently on certain 7 
fronds of Osmunda regalis, viz. the occurrence of synanagia: one of these 
is represented in Fig. 296 dis: each half shows, except on the side where 
it is coherent with the other, the ordinary sporangial structure. Such 
developments are no very surprising consequence of the origin of two 
sporangia close side by side, with a deeply seated central cell; normally, 
each develops separately, but the obliteration of the groove between them 
would give a synangial structure, and clearly this is not a very great 
modification. But what interpretation is to be put upon such cases, 
whether they may be counted as retrogressions or the reverse, must for 
the present be left open. 
The more robust and massive origin and structure of the sporangia. 
of the Osmundaceae than of ordinary Leptosporangiate Ferns is to be 
connected with the greater number of spore-mother-cells produced, and 
consequently the greater potential output of spores. In Osmunda Russow 
long ago estimated the number of spores in the single sporangium as over 
500, and assumes therefore the number of spore-mother-cells to be 128.1 
1 Vergi. Unters., p. 87. 
