UNDERWOOD: REVIEW OF THE GENERA OF FERNS 251 
with sufficient directness as to be recognizable, or (6) Based on 
some species which is described for the first time at the establish- 
ment of the genus itself. Generic names founded with no hint of 
a species on which they can rest as a type will not be considered 
as holding any priority rights against genera capable of being 
anchored to definite type species. * 
3. For each genus established the first named species will be 
regarded as the type, and to insure stability of nomenclature, and to 
prevent the shifting of generic names to groups of species wholly 
unlike those for which the original author founded the name, it 
will be regarded as essential that the type species and the generic 
name shall be inseparable and shall be maintained or reduced to 
synonymy according to the necessities of the case. 
In the application of this rule the following exceptional cases 
must be noted: (a) In the genera of Linnaeus’ Species Plantarum 
of 1753, the fact must be borne in mind that while the genera date 
from 1753, they did not all originate with Species Plantarum or 
even with Linnaeus. These Linnaean genera must then be traced 
to their type species wherever they originated. In case the 
original generic name was used in another sense than that in which 
it was adopted by Linnaeus, the type of the genus zz the Lin- 
nacan sense must be determined wherever it was first used; and 
(4) In genera established with a definite statement of the type on 
which the author founded the genus, this type must have prefer- 
ence even tho in a subsequent list of species it is not the first 
enumerated. § 
* Previous discussion of subject of generic me: has appeared hitherto by 
Y F. Cook in Bull. Torrey Bot. Club, 22: 431—434. O. 1895, and in Science, II. 
.8: 186—190, 12 Au. ied 513-516, 14 O. 1898, = B. L. Robinson has also par- 
ps stated the problem in Bot. Gaz, 25: 437-445. 1898. 
Je. 189 
f That this has often happened in the past can be shown in many cases, e. g., 
ANephrodium. "This genus, the 2 Runcphrocium of Synopsis Filicum, has been limited 
to species of Asfidicae with reniform indusia and veinlets of contiguous groups united, 
"whereas the genus. Ne ephrodium : as established by Richard in Michaux’s Flora Bor. 
mot contain a single species of the group to which it has been thus limited ! 
s, e. g., Trichomanes, whose type was Asplenium Trichomanes as known to 
Linnaeus. Hemionitis and Lonchites are other examples; see below under 175. 
2 While no general ruling has ever been made on this question, many botanists pro- 
fess to use the so-called ** method of residues"; but even here, no systematic attempt has 
ever been made to carry out this principle, ind many known exceptions occur in prac- 
