VIII 



MARATTIALES 



295 



this condition is due to a coalescence of originally separate cells, 

 or what seems more likely, arises simply from nuclear division 

 in the young tapetal cells, without the formation of cell walls, 

 was not decided. The young spore tetrads, at this time, are 

 embedded in an apparently structureless mucilaginous matter, 

 which stains uniformly with Bismarck-brown. This mucilage 

 apparently is secreted by the tapetal cells for the nourishment 

 of the spores. 



Bower (17) has recently made a very complete study of the 

 development of the sporangium in all the genera except 



A. 



B. 



Fig. 166. — A, Transverse section of three synangia of Danaa alata, X15; B, horizontal 

 section of a synangium, showing the numerous loculi, X15; C, vertical; D, hori- 

 zontal section of a synangium of Kaulfiissia cesculifolia, XiS- (C> D. after 

 Bower.) 



Archangiopteris. He finds in all of them that the sporogenous 

 tissue of each sporangium (or loculus), can usually be traced 

 to a single mother-cell, although .there may be exceptions to this 



rule. 



In all cases the tapetum arises from the tissue adjacent to 

 the archesporium, and not from the outer cells of the sporog- 

 enous complex. In this respect the Marattiacese resemble more 

 nearly Helminthostachys or Botrychimn than they do Ophio- 



glossiim. 



In Dancra and Kmilfitssia there is no mechanical tissue rep- 

 resenting an annulus. The dehiscence is accomplished by a 



