In the American species the stone layer is usually very thin ? 

 but quite distinctly indurated, — silieifiearion in reality shoving 

 density of walls better than other types of conservation. The 

 ne superficies appear to have had a very thin 

 tangential wall, even subject to collapse, so as to sometimes 

 simulate a series of flattened hexagonal cups with very heavy 

 bottoms and sides thinning a little toward the top. In the 

 Freaen - ecimen it is the cell contents rather than the cell 

 walk that appear to have ~ taken the stain/ 7 and this prodoees 

 a certain appearance as of a decidedly fleshy tissue especially 

 in the outer cells bearing the - Hene ffif." Bat what die true 

 nature of the tissue is cannot rest in any doubt : for it is as we 

 sec ffaen highly indurated in the American specimens, and 

 nearly as much so in the 1-1 - I Wf^L: srecinien. While, as I 

 fully believe, the slightly less indurated condition seen in 

 1 U M arte Txaetly paralleled by the middle 



layer in Pinu <*a. In considering any mineralized 



- occasionally necessary to note that the 

 staining could exhi it the cell contents rather than the cell 

 walls, which might Anally fail of dear outline and be replaced 

 '---.:.- - .- '.--• t '. 



1m t J — It inner fleshy layer or ti— m re _: :»n lying 



between the stc>ny region and the nueellar sack is of about - h e 

 same breadth as the si n i egion in all the Cyeadeoidea?. bean _ 

 in mind the redueed testa in the American . ~lere 



both tLr- its are ~ sy thin. In no instance. how-~er. is the 

 inner flesh Bstinetly e_:7-^ - a conserved in the lateral regie:: : 

 the aeed feo permit camera lueida drawi:^- _:_ :: :- 



— ;:r ^:. vL- -1: — 1^ :: ::ler :!:.' :.- :. :\-:r~ ::liie ":::: i 

 soft < \--ze. Bat in :_e y : rmandy seeds preservation 



is ften diagrammatieally dear near : oase where this 



zone is from three to ab - a cells deep, with the oota 



cells elongate, and flat squarish to rounded forms on the interior. 

 I e nature and success: m »i the :r-~ral elements is clearly 

 illustrated in the transverse — . :; - through the radio- 

 - _metrie seed base shown in fig. 1. where the relation to the 

 central buri e is still dear and the layers appear in their 

 initial and sun] Soma. That the ancestral endosareal 



bundle syste is ::>t entirely eliminated is probable, though 

 final determine- : this point still depends on favorably 



located rib-angle section- : e :: :eptionally conserve 



Homologies of the teed coat* in Cycadcoidea. 



In my American Fossil Cyeads 174 die - :aken 



that Xageme&omm presents the nearest known analogy to the 

 seed wall structure nl ni#>id*fi of any form existing m 



