i8 
into use too manj obsolete combinations, the undersigned would approve 
the appointment of a committee by the International Botanical Congress 
to draw up a list to include not over 400 to 500 well known and widely 
distributed species, such as Ifomoea Pes-caprae, Ahutilon Avicennae, Alntis 
tncana, etc., the current names of which, although not in accord with priority 
under the genus, might by special agreement be allowed to stand. 
3. Subspecific, Varietal, and Formal Names. 
Subspecies, variety, and form are not sharply definable or mutually 
exclusive categories. It is therefore better that, although their sep- 
arate rank is maintained for classificatory purposes, their names 
should be regarded as forming a single nomenclatorial class. In 
determining which of several subspecific, varietal, or formal names is 
to be used in a particular case two courses are open; namely, (i) To 
choose the first name employed for the plant in question in whatever 
rank and under whatever genus published, or ( 2 ) To select the first 
name which was applied to the given plant as a subspecies, variety, 
or form under the cor7'ect generic a?id specific ?iame. The second of 
these practices is by far the simplest and therefore the best. If, as 
is believed by the undersigned, the name of a species should be the 
first generically correct binomial, it is a natural and logical corollary 
that the name of a subspecies, variety, or form should be the first 
generically and specifically correct tri- (or quadri-) nomial. The 
great practicality and convenience of this rule will scarcely be appre- 
ciated by those who have not tried it. 
4. Homonyms. 
It is obvious that the later of two homonyms should be promptly 
abandoned in case the earlier one is a valid name. If on the other 
hand the earlier of two homonyms is universally regarded as invalid, 
its existence in synonymy forms no adequate reason for disturbing the 
second. When, however, in the course of investigation the earlier 
homonym is revived as a valid name, the later one must give place and 
the second plant or group of plants must be renamed. A valid name 
is one in use. It is a mistake to regard as valid a vague or poorly 
characterized name merely because it cannot be definitely entered in 
synonymy. For instance, there is no reason to abandon Seiaria of 
Beauvois because of Setaria of Acharius, the latter being a name 
