DAVALLIA. 195 



108. D. urophijllu, Wall. Cat. n. 2683 ; from the moun- 

 tains of Sylhet, 1 have not been able to find in my own, nor 

 in any other collection. 



ion. D. cordifolin, lloxb. ; " creeping, petioles smooth, 

 fronds about inches high long-cordate polished and very 

 firm oppositely ])innate, leaflets confluent falcate obtuse 

 deeply crenatc the lower pair semicordate, fructifications 

 marginal with very distinct roundish hard scaly involucres." 

 Roxh. Crypt. PL Jnd. 52. 



Hab. Mountains north of Rohilcunde, Roxburgh. " Tlie habit of this 

 species is remarkably hard and smooth." 



110. D. serra/a, Roxb.; "rachis 3-sided, fronds alternate- 

 ly pinnate, leaflets linear firm and smooth serrate, fructifica- 

 tions solitary at each serrature, involucre forming a small 

 pouch after the capsules expand." Roxb. I. c. 52. 



Hab. Prince of Wales' Island, Dr. Roxburgh. 



111. D.piiosa, Roxb. ; "fronds alternately bi- and tripin- 

 nate hairy, leaflets deeply crenate or pinnatifid, ultimate seg- 

 ments therefore semicircular somewhat crenulate and gene- 

 rally 3-flowered, involucre ciliate and separating outwards." 

 Ro.vh. I. c. p. 53. 



Hab. Eastern parts of the Delta of the Ganges, and thence introduced 

 by Dr. Buchanan to the Calcutta Botanic Garden, Roxburgh. 



112. T). trapeziformis, Roxb.; "stipes smooth nearly as 

 long as the ovate-oblong alternately bipinnate and tripinna- 

 tifid fronds (1 — 2 feet high), leaflets subtrapeziform obtuse 

 and more or less divided into rounded segments, fructifica- 

 tions generally in a single spot near the bottom of the fissures 

 of the ultimate segments, involucre opening on the anterior 

 margin forming a pouch." Roxb. I. c. p. 54. 



Hab. Malaccas, Roxburgh. " Of a soft texture, tending to be villous.'' 



(D. aiigustifoUa, Roxb. 1. c. p. 51, is D. auyustata, Wall. 

 (et nob. supra, p. 152), according to Griflith). 



D. multijloray Roxb. 1. c. p. 53, is D. parallela, Wall, (et 

 nob. supra, p. 153) according to the same authority. 



D. lon(jifolia, Roxb. 1. c. p. 52, is probably D. Emerso)ii^ 

 Hook, et Grev. supra, p. 150. 



(D. ? achill(C//olia, Wall. ; caudex ? flagelliform extremely 

 long, semiterete branched stout creeping here and there 

 woolly rooting s])inescent, fronds distant alternate from two 

 opposite sides of the caudex (.'') bipinnate, pinnic articulated 



