SOME REMARKS ON NOMENCLATURE 199 



montanus plant cited has tlie habit of Knautia silvatica or 

 K. inteqrifolia. In the National Herbarium is a type labelled 

 " Helvetia," correctly determined by Banks as /S'. silvatica '* H. L." — 

 i. e., compared with Herb. Linn. 



Schinz and Thellung conclude by offering their comments upon 

 the questions raised by Sprague in this Journal (1921, 129; 1922, 

 129). This seems the time to raise a few points 1 would add to the 

 discussions. 



A. The code should be as simple as possible. The method of 

 precision by means of a type — i. 6?., the individual of a group which 

 must retain the name whatever changes the group may suffer — would 

 lead to simplification and has been found to work well in Zoology. 

 The acceptance of this really valuable method might form a basis 

 of agreement with those Americans who refuse to accept the Vienna 

 code. 



B. Divergent interpretations might be settled by a system of official 

 case-law. An example of every known class of case could be appended 

 to the last rule on which it depends : if numbered serially we could 

 then choose the one similar to that under discussion and say " by 



[Art. Di] Case n the correct name is X Y ," and so save 



much print and argument. Cases could be dealt with by a committee, 

 and would form a fixed basis for discussion of the various principles 

 involved in the whole code. Some method of eliminating divergence 

 of interpretation is necessary, and this would seem to me to be the 

 best and simplest to work. 



C. A third suggestion would involve a change of rule, but not of 

 "principle." To state in Art. 4 that the first essential point is 

 " to aim at fixity of names," and in Art. 2 that " the principles aie 

 the foundation of the rules," and then to frame rules which allow tl e 

 possibility of a single plant having one name as a species, another 

 as a subspecies, a third as a varietj^ a fourth as a subvariety, etc., 

 seems inconsistent. It destroys "fixity of name" and involves a con- 

 tinual change of type, whereas it is only by means of the method of 

 types that any precise fixity can be obtained. The rules seem in this 

 respect to be antagonistic to progress. In Art. 48 and 49 there 

 should be no question of taxonomic rank. One name for one natural 

 group should be the aim. This was the practice of the DeCandoUean 

 code, and, unless the present method was made as an act of giace in 

 return for the giving up of the " Kew Kule," it is difficult to see why 

 such a reactionary idea was introduced. Every practical nomenclator 

 will have met dozens of cases where recognised names disappear under 

 this rule ; not only so, but possibly not a tithe of the changes have 

 yet been made. When its effects are contemplated, it is not surprising 

 that some persons will have nothing to do with a code which includes 

 it and ignores the method of types. 



I would further suggest that the method of alteration of code 

 should be by printing the proposed changes after general debate at 

 one conference, passing them finally only at the next conference, by 

 which time their effects and value (or harm) will have become mani- 

 fest. 



