ON GENERA AND SPECIES. 61 



5. ON THE GENERA AND THEIR CLASSIFICATION, AS FOUNDED ON 

 THEIR MODES OF GROWTH AND HABIT. 



In the preceding pages I have given a general view of 

 the principal writers on Ferns, and their systems of classi- 

 fication. I will now proceed to explain the reasons that 

 have induced me to attempt to draw up a natural arrange- 

 ment of Ferns, reviewing as I proceed the data upon 

 which the genera are founded. 



In my first general arrangement in 1841, 1 characterised 

 the whole of the circinate Ferns under divisions and tribes, 

 as shown in the abstract of my arrangement, given at 

 page 33, the characters of which are in general obvious, so 

 that any Fern with perfect fructification can readily be 

 referred to its proper tribe. A system of classification so 

 based is therefore practically useful, but it must be viewed 

 as purely artificial, for, although many of the genera in 

 any one tribe are naturally allied in general habit, yet 

 others of totally different habit, but possessing the technical 

 character of the tribe, are found associated with them ; 

 while on the other hand genera closely allied in natural 

 habit and organisation are placed in separate tribes. For 

 example, the genus Goniopteris has naked sori, while its 

 congener Nephrodium has indusiate sori, they are conse- 

 quently placed apart in any arrangement which is based on 

 the absence or presence of indusia. In both these genera 

 the species are numerous, and in many cases are so alike 

 that in the absence of the indusium, specimens of Nephro- 

 dium are referred to Goniopteris. The same applies to the 

 extensive genera Lastrea and Phegopteris, both of which 

 contain closely allied species. 



When the indusium is absent in normal indusiate genera, 

 it is in general impossible, even by the most careful micro- 



