November 22, 1907] 



SCIENCE 



721 



genus, or the type of a previous monotypic 

 genus, or a species some earlier reviser has 

 properly chosen as the type of some other 

 genus, rules a to d clearly show that his work 

 must be construed as void. Evidently an 

 earlier monotypic genus can not be canceled 

 by the act of some blundering reviser who 

 chances to seize upon its only species as the 

 type of some other genus; nor can a genus 

 with a " type by subsequent designation " be 

 canceled because its type was later made the 

 type of another genus. This would seemingly 

 all go without saying were it not that some 

 systematists assume that the designation of a 

 type by a first reviser is sacrosanct and must 

 Stand regardless of any other considerations. 



This emphatic reaffirmation of the principle 

 of " type by subsequent designation " is ex- 

 ceedingly gratifying. Yet, for reasons in part 

 already stated, it is to be regretted that the 

 International Commission did not define the 

 manner of its application. This doubtless did 

 not seem necessary; but there is apparently 

 nothing so uncertain as the point of view 

 from which any problem in nomenclature may 

 be approached. 



The great utility of the " type by subse- 

 quent designation " rule as an aid in estab- 

 lishing genotypes is not at first apparent; and 

 in recent years it appears to have been to a 

 great extent overlooked, it having been re- 

 garded by many as vague and illusory, and 

 difiicult to apply with certainty and precision. 

 That the principle was formerly respected and 

 extensively and effectively employed is evident 

 from a study of nomenclatural progress dur- 

 ing the last half century. My recent investi- 

 gations in an attempt to show how the types 

 of the genera of North American birds were 

 determined,^ to which Mr. Stone has recently 

 directed attention,' resulted in disclosing the 

 extent to which the currently accepted types 

 of polytypic genera in ornithology have been 

 fixed by " subsequent designation." 



As Mr. Stone has well said (Z. c.) : 



' " The Types of the North American Grenera of 

 Birds," Bull. Amer. Mus. 'Nat. History, Vol. 

 XXIII., pp. 279-384, April 15, 1907. 



= Science, N. S., Vol. XXVI., pp. 444-446, Oc- 

 tober 4, 1907. 



Much of the chaos in generic nomenclature 

 which has become intolerable to the systematist 

 of to-day has been brought about by the failure 

 of many writers to explain by what process they 

 have determined the types of old polytypic genera. 

 Had they been more explicit upon this subject, we 

 should have been able long ago to see the weak- 

 nesses in our codes and should have abandoned 

 methods which were neither definite nor final in 

 their operations. 



In fact, it is only about thirty years since 

 it became the practise for even monographers 

 to give types for genera founded by previous 

 authors except sporadically, and rarely has the 

 method of their determination been stated, 

 except in the case of types determined by 

 elimination, beginning with the A. O. U. 

 Nomenclature Committee in 1886. There is 

 merely the bare statement that the type of a 

 genus is a certain species. Usually it is neces- 

 sary to trace back the literature to ascertain 

 whether the genus was originally monotypic, 

 or whether the type was designated by the 

 formder, or determined in some other way. 

 Nor can this be fully shown until, in addition 

 to giving the author, date and place of descrip- 

 tion, is also given the original constitution of 

 the polytypic genera, with a list of the species 

 and their final generic disposition. 



My purpose in preparing the paper above 

 cited was to ascertain for my own satisfaction 

 two things: (1) whether it was true, as al- 

 leged, that no two investigators could reach 

 practically the same results in type determina- 

 tion by the method of so-called elimination; 

 (2) to determine the relative number of 

 changes necessary in the generic names of 

 North American birds by elimination and by 

 the first species rule. An entirely independ- 

 ent, or de novo, application of elimination 

 resulted in only three changes chargeable ex- 

 clusively to elimination, equal to about three 

 fourths of one per cent, of the total number 

 of genera and subgenera involved; twenty 

 would be necessary from the enforcement of 

 the first species rule, with all the Linnwan 

 genera excluded, eighteen of which have re- 

 eeired the approval of the A. O. U. Commit- 

 tee, acting tentatively under the first species 

 rule. 



