279 



cient. The writer of this, furnished with a vast collec- 

 tion of Ferns in the Linnaean herbarium, and from that 

 of Sir Joseph Banks, first suggested an additional prin- 

 ciple of arrangement, derived from the form and inser- 

 tion of the membranous cover, or involucrum, and espe- 

 cially from the direction in which that part bursts, or 

 separates from the frond, when arrived at maturity ; 

 whether, if lateral, at the side towards the margin of the 

 frond, or of its segments, or towards the rib or vein ; or, 

 if tei'minal, towards the extremity, or contrariwise. This 

 principle is found to produce very certain distinctions, 

 and to establish the most natural genera. All subsequent 

 writers on Ferns have adopted it. First, Dr. Swartz, in 

 his Syjiopsis Filicum, considers the part in question as of 

 eminent importance in defining the genera, and has es- 

 tablished several new ones on characters taken there- 

 from ; bestowing liberal commendation, in the 5th page 

 of his preface, on his friend the inventor of this method. 

 Willdenow in his Species Plantarum, vol. v., follows it 

 without a word of acknowledgement, as does Dr. Kurt 

 Sprengel in a rather superficial work, scarcely worthy of 

 its able author, translated into English by Mr. Konig, 

 under the title of An Introduction to the Study of Crypto- 

 gamous Plants, 1807; the figures of which are quoted in 

 the following pages, and will be found very useful in the 

 illustration of other vegetables of this class. But what- 

 ever deficiency of candour or knowledge may exist in 

 other writers, Mr. Brown, in establishing the curious 

 genus Woodsia, has done ample justice to his friend's 

 claims, which no one was more competent to appreciate. 

 See Trans, of Linn. Soc. w. 11. 170. 

 The roots of Dorsiferous Ferns are perennial, either tube- 

 rous or creeping, scaly, often parasitical, with crooked 

 siont radicles. P/a?i^5 mostly herbaceous, natives of shady 

 or damp situations, in almost all climates, either ever- 

 green or deciduous ; those of tropical countries some- 

 times arborescent, and occasionally spinous ; the pubes- 

 cence of all, in general, scaly rather than hairy. ivoMc? with 

 a simple or alternately branching stalk ; the leaf either 

 firm or more rarely membx'anous, ribbed, veiny, palest 

 in some degree at the back ; either simple or variously 

 pinnate ; undivided, pinnatifid, or lobed ; the divisions 

 mostly alternate ; entire or serrated ; bearing i\\Qfructi- 

 €cation at the back, very seldom at the edges, as above 



