v ; PREFACE. 



characters of the earlier authors; in whose day, the position and form 

 of the sori, with the presence or absence, or the form and direction of 

 the indusium, comprised the only points of generic distinction. Such 

 characters prove to be by no means satisfactory, as they frequently 

 brought species together which in habit and structure have little or 

 no affinity with each other. 



By the publication, in 1836, of Dr. Presl's " Tentamen Pteridogra- 

 pMce," and by Mr. John Smith, in 1840, of "An Arrangement and 

 Definition of the Genera of Ferns, &c, &c." a wider field has been laid 

 open, by the promulgation of new generic characters, derived from the 

 position and form of the sori, in conjunction with the nature of the 

 venation (the importance of which was long ago pointed out by Mr. 

 Brown) ; to which Presl has added that of the form and number of 

 the vascular bundles in the stipe. These characters, especially those 

 taken from the subordinate modifications of the venation, have re- 

 sulted in a great multiplication of genera ; many of which are based 

 upon single and trivial points of character, and might with more pro- 

 priety be viewed as sections or subgenera. The latter conviction has 

 been forced upon me during the progress of this work. But, under 

 the circumstances in which 1 was placed, I could not pretend to form 

 and act upon an independent judgment in respect to the generic 

 arrangement of Ferns in general. It was, therefore, not only con- 

 venient, but necessary, for me to follow some published system as a 

 whole, although I might entertain a confident opinion that the genera 

 were, in many cases, multiplied beyond what a sound judgment would 

 approve. Of the arrangements proposed, that of Dr. Presl is the 

 most elaborate, and accompanied by the most complete catalogue of 

 species and tables of illustrations; and it has been of the greatest use 

 in facilitating the study of Ferns. The generic characters, however, 

 which this author derives from the number of the bundles of vessels 

 in the stipe are neither satisfactory nor readily available, varying as 

 these bundles do in species of the same genus, both in form and num- 

 ber; yet specific characters of this kind might be employed to some 

 advantage, particularly in the examination of living Ferns. But in the 



