1898] THE ROCHESTER NOMENCLATURE 439 
agreement of botanists. Itis not my purpose here to discuss the 
ative merits of the absolute and usage systems, for that has 
already been done ad nauseam. I merely wish to show that the 
Rochester nomenclature corresponds to neither of these sys- 
tems; that it falls between them; and that while claiming to 
fest upon a firm basis of priority, it derives many of its principal 
Within the last few years two kinds of priority have been 
ognized, that of time of publication and that of relative 
definite and almost as necessary to an absolute system of reform 
as the more generally recognized priority of time. Both have 
been acknowledged principles in the Rochester reform, but the 
teformers in their application of the “priority of place’’ have 
en neither thorough nor consistent. While they have felt it 
necessary to discard many well-established names on account of 
$ principle, they have failed to apply it when determining 
ich of several species is to be regarded as the type of a 
By way of illustration we may consider the Linnzan 
enus Erysimum, which, according to the theory of the Roches- 
ter code, dates from its treatment in the first edition of the 
ut strangely enough our reformers, while professing to follow 
ority as the ‘fundamental principle” of nomenclature, 
lected not the first but the last species of the Linnean genus 
fetain the name Erysimum. In other words, they have here 
abandoned the much-extolled principle of priority and have 
adopted one of usage. They have taken Erysimum cheiranthowdes 
the true type of the genus, not because it was the first 
‘Species to bear that name, but because it was the species which 
had been so treated by subsequent usage. Nor is Erysimum an 
