Apkil 22, 



3.] 



SCIENCE. 



573 



ience according to the presence or absence of 

 some unimportant structure. Presl, who was 

 one of the first to recognize natural divisions 

 among the ferns, gave us 230 genera ; John 

 Smith, with the advantage of the Kew collection 

 and the largest number of species anywhere in 

 cultivation, only reduced this number to 220; 

 while Fee, the illustrious French pteridologist, 

 recognized 181, and Moore a little later 177. 

 The Kew authorities recognize only 78 includ- 

 ing recent additions. 



The system followed in the present work can 

 be best judged perhaps by the following outline 

 of its main features : 



LEPTOSPOKANGIAT^. 



{Polyangia.) 

 Hymenophyllacete. 

 Polypodiacese. 



Acrostichese. 



Vittariese. 



Gymnogrammese. 



Polypodiese. 



Pteridese. 



Aspleniaceffi {sic). 



Aspidiacese (sic). 



Davalliacese {sic). 

 Cyatheaceffi. 

 Osmundaeese. 



{Oligangia.) 

 Matoniacese. 

 Gleicheniacese. 



{Monangia.) 

 Schizseaccce. 

 ParkeriaceEe. 



EUSPORANGIAT^. 



Marattiacefe. 

 •Ophioglossaceae. 



Concerning the arrangement of families there 

 would probably be little difference of opinion 

 except that from an evolutionary standpoint 

 the order should be inverted and the position 

 of the Hymenophyllacefe would be called in 

 question. The tribes of the Polypodiaceae will 

 permit more diversity of opinion. Among the 

 good points to be noted are the removal of 

 Notholsena (wrongly printed Nothochlsena) to the 

 Pteridese, where it stands next to its close ally, 

 Cheilanthes ; the removal of Lindsaya, Nephrol- 

 epis and Loxsoma, to the Davalliese ; also the 



formation of distinct families for the aberrant 

 Matonia and Ceratopteris. The separation of the 

 unique Platycerium from the Acrosticheie is well 

 timed, but it finds a resting place just as unsat- 

 isfactory next to Polypodium and always will be 

 a migi-ant until it is placed in a distinct family of 

 which it is worthy on account of its unique 

 characters. 



The divisions of genera are interesting, but 

 exceedingly unequal. Acrosiichum as recog- 

 nized by Baker here appears under seven or 

 eight genera, but the equally composite Gymno- 

 gramine is grouped under only three generic 

 names, though some of its species are relegated 

 to Polypodium and Phegopteris. Hemidiciyon, 

 Ceierach, Diplazium and Athyrium (the latter in- 

 cluding A. filix- foemina) are separated from 

 Asplenium. while the equally distinct Thamnop- 

 teris {A. nidus) and Darea are still left in the 

 genus, and the more distinct Camptosorus and 

 Schaffneria are still left in the same genus as 

 Scolopendrium. Struthiopteris is verj' properly 

 separated from Onoclea, and Cibotium and Denn- 

 stedtia from Dicksonia, though in each case 

 there are complications of nomenclature that 

 will demand a later settlement. Both Polys- 

 tichum and Nephrodium are united under As- 

 pidium and Fadyenia is also included, contrary 

 even to the conservative Kew practice. We are 

 still very far from a natural segregation of the 

 genera of ferns. 



The recognition of species, especially those 

 of American origin, is exceedingly faulty. The 

 combination of Notholsena Candida, N. Hookeri 

 and jV. cretacea into one species is no less noto- 

 rious than the reduction of Aspidium Goldieanum 

 and A. marginale to varieties of A. filix-Hymas. 

 Quite a number of our American species appear 

 under new names, which will be a new source 

 of grief to conservative botanists, who are 

 troubled because names will change with the 

 advance of investigation. Among these not 

 already mentioned we note : 



Chrysodium aureum instead of Acrosiichum 

 aureuvi. 



Neurodiiim lanceolatum instead of Tsenitis 

 lanceolata. 



Blechnum spicant instead of Lomaria spicant. 



Dennstedtia punctiloba instead of Dicksonia 

 punctiloba. 



