64 TklKSOZOIC FLORAS OF UNITED STATES. 



They are now slightly {'ontracted at the attachment of their bases, 

 but are still decui'rent to form a wing. When the incision of the mar- 

 gins is least, they are undulate or crenate. With deeper incision they 

 are serrately dentate or lobed, the teeth and l()l)e8 l)eing a(;ute and having 

 more or less of the character of the simple pinnules into which the lobes, 

 with increasing deptli of incision, finally pass. These incised pinnules 

 are attached at an angle of about 45°. The fertile pinnules show a 

 greater tendency to obtuseness, the teeth and lobes being often less 

 sharply serrate. 



The sori are placed much as they are in Dkksonia oregonensis. 

 In the simple pinna? there is only one sorus, which is placed within the 

 margin, on the end of the unforked branch of the nerves, in the anterior 

 basal part of the pinnule. The great size of the sorus causes it to fill 

 much of the basal part of the pinnule on tlie upper side of the parent 

 nerve. In the incised pinnules there is one sorus for each tooth or 

 lobe, and they are placed on the ends of the lateral nerves. The sori 

 are without indusium, and under the pressure to which they have Ijeen 

 subjected appear as flat, rounded patches, with a granulation, which 

 is apparently due to the naked capsules. The granules are grouped 

 around a central point, which seems to have been the receptacle. When, 

 however, the sori are seen with the upper surface of the pinnules pre- 

 sented uppermost, they are a good deal disguised and their true natiu'e 

 could not be made out. They are then smaller and appear as rounded 

 prominences. The fructification seems near enough to that of Poly- 

 podium to justify the placing of the plant in that genus. 



Owing to the considerable number of specimens that were obtained, 

 a good idea can be formed of the plant. It must have been of large 

 size, with widespread pinnae, for in their great length they contrast 

 strongly with those of Dicksonia oregonensis. As in the case of this 

 last-named plant, by putting together fragments from different parts 

 of tlie compound pinna a partial restoration may be made. It is a much 

 more elegant fern than D. oregonensis. 



PI. VIII, Fig. 12, gives fragments of two penultimate pinnse, which 

 carry ultimate pinna? and sterile simple pinnules. The penultimate pinnae 

 were evidently once united to a pinna of superior order. The fragments 

 come from near the ends of the penultimate pinna? or else from high up on 



