XXX] 



DENNSTAEDTIINAE 



271 



is not wholl}' interrupted by its insertion; in such cases it is actually con- 

 tinuous at the base, as it is in the more primitive sporangia with their 

 conspicuously oblique ring. But not uncommonly the ring is actually inter- 

 rupted at the stalk. In fact these sporangia show only faint traces of 

 obliquity of the annulus, and these inconstantly. The annulus is almost 

 vertical, and its dehiscence transverse. Sections of the young sporangium 

 show that the number of spore-mother-cells in each sporangium is variable: 

 eight, twelve, or sixteen have been observed. Countings of the mature spores 

 have shown that though sometimes the full number of 64 may be produced, 

 the tale is frequently less. The orientation of the sporangia is not constant, 

 but perhaps this may be due in part to twisting of the long stalk. 

 A C D 



Fig. 539. A, Dennstaedtia apiifolia Hook. Sorus showing basipetal succession 

 of sporangia: zW = indusium. C, dehiscent sporangium of the same, showing 

 very slightly oblique annulus. B, Demistaedtia ntbiginosa Kaulf.: sorus in 

 vertical section showing that it has been at first basipetal, but a mixed 

 character has supervened. Z>, dehiscent sporangium of the same, seen from 

 the base, showing that the annulus stops short on either side of the insertion 

 of the stalk \st). (All x loo.) 



All these features indicate for Dennstaedtia an advance upon the marginal 

 types so far described. But there is another still more important for com- 

 parison. Deviations from the basipetal succession of the sporangia have 

 been observed in D. davallioides (Br.) Moore, and in D. rnbiginosa (Kaulf) 

 Moore. Even in D. apiifolia isolated cases occur of a sporangium near to 

 the apex of the receptacle in a less developed state than those below. In 

 D, davallioides such cases are fairly common ; but in D. rnbiginosa the sorus 

 bears upon the shortened receptacle sporangia without definite order of 

 succession (Fig. 539, E). The sorus has, it is true, some signs of a basipetal 

 succession at first, but later it takes clearly the "mixed" character, with 

 younger sporangia interspersed irregularly between those that are older: 

 and correlated with this we find the receptacle short, but wide. Moreover, 

 the annulus is here vertical, and definitely interrupted at the stalk (I^'ig. 



