468 



SCIENCE 



[N. S. Vol. XXV. No. 638 



choice and personal favoritism. It has, more- 

 over, shown that no rules or laws can be made 

 binding by mere agreement. They must in 

 the very nature of things represent the best 

 possible adjustment, else later generations will 

 cast them aside. Hence laws must be simple, 

 clear, and as far as may be, free from excep- 

 tions. 



As part of the nature of things, the law of 

 priority has forced its way to acceptance. In 

 like fashion the law of unchanging spelling 

 and the rule that nouns spelled differently are 

 different words, regardless of etymology, must 

 become universal. 



The chief point of variance now left is this : 

 A genus is known by its type. In case an 

 author of a genus names several species, but 

 fails to designate one as type, how shall the 

 actual type be determined? 



In this case we have two possible methods, 

 both resting on a logical basis. I do not 

 count among these the rule of elimination, 

 which in my judgment has never been defined 

 in workable fashion, and, however defined, will 

 never meet with general adoption. 



These two methods are, in brief, that of the 

 ' first reviser,' and that of the ' first species.' 



In the first method, when a genus is left 

 without type indicated, the genus rests with 

 the first of the original species which any 

 subsequent author may select as type. In 

 the second method, the type, unless otherwise 

 indicated, is the first species named under the 

 head of the genus in question, by its original 

 author. 



The logical basis in the first case is this : 

 No generic diagnosis is complete until a type 

 species has been indicated. If the original 

 author neglects this, the first of his succes- 

 sors who does it, completes the generic diag- 

 nosis, and the generic name in question must 

 stand or fall with the species selected. The 

 rule of elimination may be a handy device for 

 the use of the first reviser, and as such it has 

 been justly commended ; but he does as he 

 pleases, and from his decision, the type once 

 chosen from among the original species, there 

 is no appeal. That is, we have no appeal, 

 unless we find an earlier ' first reviser,' in 



which case the act of the later one is null 

 and void. This rule of the first reviser is 

 probably more generally recognized than any 

 other in systematic zoology and botany. It 

 has been lately strongly supported by Dr. 

 Charles Wardell Stiles. Objections to it are 

 these : Often the first reviser overlooks the fact 

 that a type was virtually or even actually in- 

 dicated by the original author. This makes it 

 necessary to reverse many time-honored deci- 

 sions, in which the work of the reviser ia 

 better known than that of the original author. 

 More often the first reviser fails to make his 

 own position clear. Whether he has actually 

 chosen a type or merely used a species of 

 illustration is often a matter of doubt. Still 

 worse, after we have followed the first reviser, 

 we find that we had overlooked a still earlier 

 one. As a result of this, no name of this sort 

 is safe until all the returns are in and all the 

 work of obscure authors has been examined. 

 This is bad enough when the priority of names 

 is in question. It is a trial to do it in the 

 interest of the meaning of names as well. 



The first species method rests on this prin- 

 ciple; the application of a generic name 

 should depend solely on data furnished by the 

 author himself. If he fails to indicate a type 

 species, either directly or in some one of the 

 recognized methods, then the first species he 

 mentions must be considered as type. If he 

 mentions no species, directly or by implica- 

 tion, the genus has no existence. This rule 

 has been lately adopted by the American 

 Ornithologists' Union and a strong argument 

 for it has been made in this journal by Mr. 

 Witmer Stone. 



There is no injustice in this rule. Its ap- 

 plication rarely if ever admits of serious ques- 

 tion. It does not involve any bootless investi- 

 gation of the meaning or intention of subse- 

 quent writers. It would involve probably less 

 change of accepted nomenclature than any 

 other rule that can be framed. The first 

 reviser, in fact, has usually chosen as type the 

 first species named by the original author. 

 The French school of zoologists, Lacepede, 

 Cuvier and their successors as a rule have 

 placed first as chef de file the typical species 



