XIII] THE WALL OF THE SPORANGIUM 249 



with the mixed type of sorus. Finally, many of the advanced Leptosporan- 

 giates have very long stalks, often consisting of only a single row of cells 

 (Figs. 239, 240). Here the stalk lengthens rapidly as the sporangium ripens, a 

 condition favourable for mutual protection of the sporangia when young and 

 liberation of the spores when mature in sori where the receptacle is flat and 

 crowded. Thics speaking generally, massive, sJiort stalks may be held as pj'imi- 

 tive, elongated and thin stalks as derivative and specialised in character. 



The Wall of the Sporangium, and the Annulus 

 The Wall of the sporangium in Ferns varies greatly both in thickness 

 and in mechanical construction. In most of the Simplices it consists of 

 several layers of cells. This is seen in the fossil Stauropteris where the adult 

 wall has some four layers, of which the outermost is strongly differentiated. 

 The wall of Helminthostachys, and of Botrychium daiicifolium, is similarly 

 constructed (Fig. 242), as is also that of Angiopteris. But in the Lepto- 

 sporangiate Ferns it consists of only a single layer of cells. Occasional 

 tabular cells may, however, be found lining the outer wall internally in 

 Osmunda, and Lygodium, which thereby confirm the position of these Ferns 

 as transitional types. We conclude from these facts that a many-layered 

 wall was primitive, and a single-layered wall a derivative state. 



The chief interest in the wall, and its special value for comparison, lie 

 in the construction of the mechanism of dehiscence by local induration of the 

 cells. In advanced cases the result appears as the annulus, and the stoininm. 

 But in the simplest no specially differentiated annulus is to be seen. This 

 is so in Stauropteris, where dehiscence is by means of a distal pore (compare 

 Fig. 197, Chapter Xll). In modern Eusporangiate Ferns, however, and in most 

 of the fossils, the dehiscence is by a slit which runs longitudinally down the 

 side of the sporangium. Angiopteris gives a good example (Fig. 244). Here 

 the wall is some three or four layers of cells in thickness : but the superficial 

 layer alone is mechanically developed, and only parts of it are indurated, the 

 rest consisting of thin-walled cells. The annulus appears as a firm vertical 

 arch, broad on either side of the sporangium, but narrow at the distal end 

 (Fig. 244, D, E, F). The slit of dehiscence runs down the side next the centre 

 of the sorus, and it is drawn open by the contraction of the peripheral side 

 of the sporangium, the annulus acting like an elastic hoop. This involves 

 a change of form possible only in a free sporangium. In Marattia, Danaea, 

 and Christensenia on the other hand; in which the sporangia are fused into 

 synangia, there is no developed annulus, and the short slit of dehiscence is 

 opened by the drying and contraction of the adjoining cells. The sporangia 

 of many Carboniferous Ferns appear to have behaved in the same way as 

 these living Ferns. Corynepteris and Etapteris had indurated vertical hoops of 

 tissue, which appear to have acted like those o{ Angiopteris (Fig. 245): the 



