THE SPORANGIUM 639 
‘diagnostic feature in Ferns, it requires special notice, and will be taken 
first. In the synangial Marattiaceae and Pecopterids the opening mechanism 
is very simple: a slit-like rupture is formed, and it gapes by drying up of 
the adjoining cells, while the firmer region of the wall stands rigid. This 
may be complicated by changes of form of the whole sorus, as in Marattia. 
But in those early forms of Ferns in which the sporangia were separate, 
there is commonly a band of mechanical tissue, composed of several rows 
of indurated cells: this band varies in extent and in position in different 
sporangial types. It has been stated by Scott,! referring to the annulus 
of certain primitive Ferns which is more than one cell wide, that 
“this was perhaps a general character of the annulate Fern-sporangia of 
Palaeozoic age: at least no clear case of a uniseriate annulus has yet been 
demonstrated.” In fact, it seems that in the Primary rocks the distinctive 
Leptosporangiate annulus was at least rare, if indeed it existed at all. 
It may be a question whether the more complex mechanism thus seen 
in Eusporangiate Ferns is really the true correlative of that in the Lepto- 
sporangiate type. A comparison of the indurated tissues in the sporangia 
of Angiopteris and Gleichenia shows certain points of essential similarity, 
though the details do not correspond. The firm resistant arch of indurated 
tissue described in Angiopteris consists of cells of a similar nature to 
those which form the annulus of Gé/eichenia; its position is in the main 
the same, though it does not stretch as a continuous hoop round the 
back of the stalk, but stops short on either side of its base. When we 
consider the similarity of the sorus in these two genera, and of their 
sporangia with the same orientation and dehiscence, the comparison of 
these two bands seems inevitable, notwithstanding that the one is usually, 
though not always, a single row, and the other abroad and ill-defined 
band. But a further important fact is that among early Ferns of reputed 
Leptosporangiate affinity the annulus is not always a single row of cells; 
this is seen in the Schizaeaceae, for Senftenbergia has an annulus of several 
rows: Zeiller has shown that living species of Lygodium may have a 
double-rowed annulus, which is an intermediate step to the type with 
a single row.2. Again, in living species of G/eichenia occasional cells of 
the annulus have been found to be divided, showing thus a reminiscence 
of a pluriseriate state. Such evidence favours the opinion that the simple 
annulus is the correlative of the pluriseriate, and that a simplification of 
its structure has accompanied the reduction in size and spore-output of 
the sporangium in the course of descent. In the Leptosporangiate Ferns 
the homogeny of the annulus seems the only view which is in accord 
with the constancy of its occurrence in plants which are so clearly related 
to one another. 
The position of the annulus and of the point of dehiscence appears to 
have undergone change in the progressive evolutionary course. In the 
1 Progressus Ret Botanicae, i., p. 184. Compare Kidston, Phil. Trans., Ser. B, 
vol. 198, p. 188. 2 Bull. Soc. Bot. de France, xliv., p. 214. 
