438 BOTANICAL GAZETTE [JUNE 
this has been done in disregard of the opinions and expressed 
wishes of a very large number of their colleagues does not con- 
cern us here, except as showing an unfortunate over-confidence 
on the part of the reformers. What they should realize, how- 
ever, is that no number of monographs or floras, published in 
accordance with the Rochester code, will establish a single 
principle or a single name, which does not appeal to future 
botanists as reasonable. Surely those who have themselves 
discarded hundreds of names which had stood unchallenged for 
nearly a century should not feel that they are establishing their 
system merely by putting it into use. The only way it can be 
established is by making it so reasonable and consistent that it 
will command general respect and approbation. 
Readily accepting the now generally admitted fact that 1753 
is the most desirable date of departure, the writer can see only 
two logical methods of codifying botanical nomenclature. 
According to the first of these modes, priority both of time and 
place must be unrestricted from the date of starting. Each 
plant must bear its earliest designation, and each name must be 
used only in its earliest signification. Such a system would 
involve a hitherto unprecedented change but is both conceivable 
and logical. The other method, while also recognizing the 
great value of priority in determining the proper names of 
plants, would seek to limit this principle by such qualifications 
as would be necessary to retain as great a part as possible of 
the current nomenclature. Inthe first or absolute system no 
exceptions can be permitted, usage may not be taken into 
account, and in fact nomenclature must be torn down to the 
point where it can be rebuilt with regularity and symmetry. The 
language of systematic botany must, in such a system, start 
almost afresh and follow unswervingly certain theoretical 
principles. In the other system, principles must also be sought 
out and followed, but here, like the rules of grammar, they 
should be based upon usage and derive their guiding power by 
stating, generalizing, and correlating usage and not by defying 
it. Either system to be effective requires a fairly general 
