328 



rOLYPODIACE^:. 



The resemblance of this fern or Prosaptia contigua and species of the 

 subgenus Cryptosorus, to which Polypodium obliquatum belongs, is very remar- 

 kable, and of a kind which no botanists ever dispute. The habit and shape 

 of fronds, and the structure of hairs, scales and rhizomes are all the same. 

 The hairs (Fig. 137-b) are very peculiar ; they consist of several elongated cells 

 arranged in a single row with very thick brownish walls. The scales are also 

 very remarkable, being composed of polygonal cells with rather thick brownish 

 walls arranged 



in one layer, 

 and beset with 

 bristles on the 

 outer surface 

 (Fig. 137-a). 

 Prosaptia, as 

 stated above, 

 closely resem- 

 bles Polypo- 

 dium decres- 

 cens. The 



scales of the 

 latter differ 

 from those of 

 the former on- 

 ly in lacking 



, . , , Fig. 137, a f, Polypodium urceolare HAY AT A ; a, a scale of rhizome ; 



b, basal portion of a hair on the stipe; c, portion of a frond, showing the 



affinities in venation ; d, portion of a puma, seen from below, showing the sori of the 



, type ; e, an earlier stage of the sorus-development ; f, much more advanced 



sori, part of involucres taken off, showing the receptacles ; g, an earlier stage 



all point to f ^ ie sorus development of Pubjpodium obliquatum; h, nearly the same stage 

 , i , . in Prosaptia celebicum BLTJME ; i, nearly the same stage in Prosaptia Emer- 



soni ; j, full grown sori of Davallia bullata ; k, very earlier stage of the sorus- 

 COngeneric, as development in the same species, showing the sori of the Humata-type. 



such a general agreement in character is one of the most important points in 

 the classification of genera, as is shown by C. CHRISTENSEN, and others. The 

 anatomical characters seen in the structure of hairs and scales arc mostly 



