226 BOTANICAL GAZETTE [ MARCH 
described without the citation of a type specimen. The type 
method is equally or even more necessary and applicable in the 
case of genera; but admitting the advisability and possibility of 
such a method the details still remain to be arranged. Cook, 
-d.c¢., has discussed the question in general ; Underwood, /. ¢., has 
formulated a set of rules for fixing types and applied them to — 
the genera of ferns; Jordan, /. c., has also recently given us his 
views regarding the matter. a 
It seems very desirable and important that this matter should 
be thoroughly investigated and discussed before any fixed plan 
is adopted for putting the method into practice. oa: 
The essential features of any rule should be simplicity, clear- 
ness, and comprehensiveness. Rules which require or give — 
opportunity for personal choice or interpretation cannot lead to 
uniform results. Provision should be made for all classes of : 
“eases involved, and this necessitates a thorough knowledge of 
all taxonomic literature. To one not familiar with the se 
en 
hall 
| ynomial spec 
o the time 
ing po 
satisfacto 
application of a generic name to the genus in which 
ul (pe or first binomial species is included.” ae 
ing point might be this would require that gener © 
- polynomialists should be interpreted by their first 
