IOO 
THE PLANT WORLD. 
unabated force with most nomenclaturists. In Sydow's mono¬ 
graphic work upon the Uredineae, the first volume of which is but 
just from the press, we find that the author changes the name 
Puccinia Distichlidis E. & E. to P. Kelseyi Syd., because it has 
been found that the host is not a species of Distichlis, as at first 
supposed, but Spartina gracilis, and the name is therefore no 
longer appropriate and will moreover give cause for confusion. 
Dr. Sydow says that in this he is quite in agreement with the 
majority of mycologists, that in fact new names have also been 
chosen by others in analogous cases. I have cited this instance to 
show that the descriptive feature of the specific name is still often 
considered of much and even of controlling importance. But it 
was early seen that instability would 1 esult if appropriateness were 
made the test in retaining or rejecting a name, and in the Paris 
code of 1867 changes based upon such a reason were, with some 
exceptions, prohibited. The tendency has been to give less and 
less weight fi> the descriptive feature of the name in determining 
correct usage, and correspondingly more weight to the appellative 
feature. I11 some of the recent attempts at writing or revising 
rules to govern botanical usage, this tendency has been carried 
to a logical conclusion, and while the author of a name is ex¬ 
pected to exercise all care in its selection, that it may be as de¬ 
scriptive, or as significant as possible, yet when once published, 
the nomenclatorial treatment of the name does not consider its 
meaning but only its right to designate a certain object, and that 
object is the species indicated by the specimen which the author 
employed to represent it. 
The idea of using a type or representative of the species, to 
which reference can be made whenever it is necessary to deter¬ 
mine the true application of the name, is of recent origin. It is , 
recognized that the type may be destroyed, lost or inaccessible, and 
in such cases the character of the type must be determined by 
the best circumstantial evidence available. But in the majority of 
cases the type specimen is in existence. By the use of types the 
original application of a name is more accurately ascertained than 
by any device heretofore suggested. 
But it is rare or never that a specimen represents all the char- 
