8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



oblong, long-acuminate, evenly denticulate-ciliate, nearly homoge- 

 neous in structure, the partition walls sclerotic, those of the smaller 

 (outer) cells paler, strongly sinuate, the cells irregular ; upper 

 surface dull green, glabrescent, a few minute, lax, filiform scales 

 evident at first along the middle ; sori polycarpous, marginal, seated 

 upon the slightly thickened ends of the oblique, once forked, pin- 

 nately arranged veins, adjacent, slightly protected by the narrowly 

 revolute margin, the extreme border undulate-repand, slightly 

 altered, delicately herbaceous, yellowish ; sporangia confluent at 

 maturity, glabrous ; spores triplanate, coarsely tuberculate. 



Type in the U. S. National Herbarium, no. 50959, collected at 

 Bafios, in the Andes of Peru, by the Wilkes Exploring Expedition ; 

 listed by Brackenridge ' as Notholaena sinuata Kaulf . Agreeing 

 with the type specimen are plants collected in the high mountains 

 above Lima, Peru, March, 1892, by W. E. Safford (no. 999). 

 Although these, being larger and more complete, have afforded the 

 principal data for the above description, the Wilkes Expedition plant 

 of Brackenridge is for other reasons selected as the nomenclatorial 

 type. 



Brackenridge's identifications of the Wilkes Expedition ferns, 

 though made under great difficulty, were in the main correct. The 

 present instance is a marked exception, the plant bearing no close 

 resemblance or relationship whatever to N. sinuata (Swartz) Kaulf. 

 Baker, 2 recognizing Brackenridge's error, assigned to his plant the 

 new name Notholaena Brackenridgei, but apparently never published 

 a description, merely listing it as a synonym under Notholaena dora- 

 dilla Colla, a Chilean species with which presumably he considered 

 it identical. Notholaena doradilla is, however, as may be at once 

 noted from the illustration, 3 exactly the plant described from Chile 

 by Kunze i as N. mollis. The Peruvian plant of Brackenridge never 

 having, been described under a valid name, the above description is 

 offered, with the assignment of Brackenridge's specimen as the 

 actual type because of its historical association. 



Notholaena Brackenridgei might with equal propriety be placed in 

 Cheilanthes, because of. its slightly thickened fertile vein-ends and 

 rudimentary indusia. It is one of a number of similarly intermediate 

 species and suggests the necessity of a modern revision of this 

 difficult and puzzling group. 



'In Wilkes, U. S. Explor. Exped. 16: 19. 1854. 

 " Loc. cit. 



3 Mem. Acad. Torino 39 : 46. pi. 73. 1836. 



4 Linnaea 9 : 54. 1834; Farrnkr. 1 : 115. pi. 53, f. 2. 1843. 



