OCTOBEB 4, 1907] 



SCIENCE 



445 



Birds " is therefore especially welcome, as it 

 throws a flood of light upon a hitherto ob- 

 scure subject and shows us the possibilities of 

 the various methods of determining genotypes. 

 The presentation in concise form of the origi- 

 nal basis upon which each genus was erected 

 will be exceedingly useful to the systematic 

 ornithologist, but Dr. Allen's conclusions with 

 regard to the types of many of the older 

 polytypic genera will hardly be accepted since 

 he has not followed any one method con- 

 sistently in his determinations ; and consist- 

 ency is absolutely essential in work of this 

 sort. He tells us that " from the standpoint 

 of elimination " his results agree with those 

 given in the A. O. U. check list in all but 

 four instances. 



A careful study of his text, however, shows 

 that he uses " elimination " in a surprisingly 

 broad sense and that his types are really vari- 

 ously determined by " elimination " ; " sub- 

 sequent designation " ; " general consent," and 

 " restriction." Any one of which used 

 throughout would have materially increased 

 the number of changes necessary in the check 

 list. Frequently two methods are applied to 

 the same case with different results and Dr. 

 Allen accepts one method in one instance and 

 another in the next. 



For example, 



Pelecanus, p. 301. 



Type designated by Gray, 1840 — onocro- 



talus. 

 Type by elimination — aqidla. 



Conurus, p. 335. 



Type designated by Gray, 1841 — leucoph- 



thalmus. 

 Type by elimination — carolinensis. 



Gray's designation is accepted by Dr. Allen 

 in the first instance and elimination in the 

 second, and, in both, his types agi'ee with the 

 cheek list. Obviously such choice is unwar- 

 ranted; one method or the other must be used 

 throughout. 



In other cases only one method is mentioned 

 while another method would give a different 

 result, viz. : 



Passerina, p. 35Y. 



Type by elimination — nivalis. 



Alca, p. 293. 



Type by elimination — torda. 



In both these cases Gray designated other 

 species as type in 1840, but his designations 

 so frequently accepted by Dr. Allen in prefer- 

 ence to elimination are here completely 

 ignored. 



Many of Dr. Allen's eliminations, moreover, 

 would yield entirely different results, were all 

 the involved genera considered. For example: 



glacialis, given as the type of Fulmarus by 

 elimination, was removed from that genus 

 as the type of Bhantistes, 1829, and 



garrulus, given as type of Ampelis by elimi- 

 nation, was removed as the type of Bom- 

 hycivora, 1815. 



And so in many other cases critical genera 

 are omitted from consideration. 



The recent action of the American Ornith- 

 ologists' Union and the Nomenclature Com- 

 mission of the International Zoological Con- 

 gress in repudiating the elimination method 

 has probably forever removed this " bone of 

 contention " from consideration. 



There are left, as stated recently by Presi- 

 dent Jordan, two methods of type fixing, either 

 of which will yield definite and final results— 

 the first species rule and type by subsequent 

 designation. By either method some fourteen 

 changes will have to be made in the genera 

 and subgenera of the A. O. IT. check list, and 

 in determining just what they are Dr. Allen's 

 paper will be a valuable aid. 



In this connection and with the idea of 

 increasing its usefulness it seems desirable 

 to call attention to some errors in quota- 

 tion, etc. — inevitable in a work of this scope. 

 The type of Daption will . be found to be 

 designated in the original publication, while 

 that of Thalassens desig-nated by Gray in 1840 

 was cantiaca, not caspia. Gray, moreover, 

 did not designate a type for Herodias in 1840, 

 and in 1841 selected garzetta, not egretta. 

 The type of Anous which he selected in 1840 

 was not Sterna niger Stephens (= S. stolida 

 L.) but S. nigra Linn. {^Sydrochelidon 

 nigra'). The genera Dendragapus, Hydranassa 

 and Oisiothorus are polytjrpic, not monotypic, 

 as indicated. 



Dr. Allen gives 415 instead of 435 as the 



