IMPROVEMENT IN THE ARRANGEMENT 01" FERNS. 361 



duly noticed. It may deserve consideration whether the figure and 

 surface of the spores may not be employed with advantage. The 

 structure of the sporangium is certainly of high importance, but gives 

 characters for larger divisions than genera. Much importance has been 

 attached by some writers to the mode of growth of the fronds, whether 

 continuous with the candex or connected with it by a sort of joint, 

 which would divide the whole assemblage into two great sections ; but 

 it appears to me that our experience of this character, in respect to 

 higher plants, is greatly against attaching much value to it, and I 

 cannot think its effect good in respect to natural aflBnities. I would 

 therefore wholly reject it. In the order of their value, I would rank 

 first characters derived from the sporangia ; then those from the sori 

 and indusium, which, with the texture of the frond and general mode 

 of growth, will abundantly determine the alliances, orders and tribes. 

 Minuter particulars respecting the indusium, venation and position of 

 the sorus on the vein, with any other good observations on structure, 

 will duly limit the genera. 



It is probable that the best determination of the tribes of ferns is 

 that of Presl, though he was doubtless in error in making Hymeno- 

 phyllacese a separate order (just as Lindley was in giving the same 

 distinction to Daneaceae), and a few modifications of his tribes may 

 perhaps be desirable, but he seems to have failed in appreciating the 

 higher divisions, and his nomenclature is very objectionable in form, 

 and demands correction. I cannot but wonder that in adopting Presl's 

 arrangement (with a great improvement in respect to higher divisions), 

 in his excellent introduction to Cryptogamic Botany, Mr. Berkeley 

 did not see the necessity for altering the terminations of the names. 

 Lindley's principle of making the names of the larger groups, called by 

 him alliances, and, though not generally received by botanists, very 

 generally natural and of great assistance to students, terminate in ales, 

 those of the families called natural orders in acese, and those of tribes 

 and subtribes in any other convenient form of Latin derivative adjec- 

 tives, is so manifestly useful and reasonable that it may justly excite 

 surprise that it is not universally adopted, and to extend the law of 

 priority to the terminations of such names is altogether preposterous. 

 Neither Presl nor Berkeley meant to maintain that the tribes of ferns 

 are of value equal to natural orders in other parts of the system ; and 

 even if they cannot see the merit of Lindley's plan, it is exceedingly 

 injudicious to set it at defiance, and seemingly attempt to cause confu- 



