510 



SCIENCE. 



[N. S. Vol. XVII. No. 430. 



investigation within the group which he called 

 Gypaetidse (i. e., the Accipitres) in certain 

 variations of the coracoid and scapula. Tak- 

 ing up this clue, the present writer puhlished, 

 in 1873-1876, a series of papers on the classi- 

 iication of the Accipitres, and in the first of 

 these* indicated a new grouping of the genera, 

 but without definition of their characters, the 

 salient feature of the new arrangement being 

 the separation of the true falcons (Palcones), 

 laughing falcons (Herpetotherese), wood fal- 

 cons (Micrastureas) and caracaras (Polybori) 

 as a subfamily (Falconinse), the remaining 

 members of the order being segregated into 

 minor groups under the subfamily heading 

 Buteoninss. This arrangement was further 

 elaborated, with slight modifications as to 

 some of the minor groups, in 1875 f and again 

 in 1876. J 



This ' new arrangement,' so radically dif- 

 ferent from any other, found little favor 

 among ornithologists, and had apparently be- 

 come forgotten; therefore, the author hopes 

 that he may be excused the .surprise and 

 gratification which he naturally feels to find 

 in Mr. Pycraft's paper, published nearly thirty 

 years later, a classification so closely identical 

 with his own that differences of nomenclature 

 constitute almost the only points of diver- 

 gence. No mention of the present writer's 

 papers of 1873-76 on the same subject being 

 made by Mr. Pycraft, it is probable they were 

 unknown to him, or at least that he never saw 

 them, a probability the more gratifying since 

 results which have been independently reached 

 by two widely separated investigators must, 

 necessarily, be sound; and now that the 'stamp 

 of authority ' has been given to the present 

 writer's long ignored arrangement, it will be 



* ' Catalogue of the Ornithological Collection in 

 the Museum of the Boston Society of Natural 

 History,' II., Falconidse. Proc. Boston Sac. Nat. 

 Hist., XVI., May 21, 1873, pp. 43-106. 



f ' Outlines of a Natural Arrangement of the 

 Falconidce,' Bull. U. S. Geol. and Oeog. Surv. Terr., 

 No. 4, second ser., June 10, 1875, pp. 225-231, 

 pis. xi.-xviii. 



t ' Studies of the American Falconidse,' Bull. 

 V. 8. Geol. and Geog. Surv. Terr., ii.. No. 2, 

 April 1, 1876, pp. 91-182. 



safe for the conservative ones . to shake off 

 their adherence to antiquated and obviously 

 unsound classifications of the group and 

 adopt the modern one. The latter, it is 

 hardly necessary to remark, is of course sus- 

 ceptible of much improvement, especially as 

 to the number and limits of the minor groups 

 (called subfamilies by Mr. Pycraft), there 

 being still many forms whose osteology has 

 not yet been studied. 



In order to show how very closely the pres- 

 ent writer's arrangement of 1873-76 coincides 

 with Mr. Pycraft's of 1902, the two are com- 

 pared in parallel columns, with a few neces- 

 sary explanatory notes: 



Ridg-way (1873-76). 

 Falconing (1873). 

 Falcones (1873). 

 Polybori (1873). 

 Micrasturese* (1873). 

 Herpetothereset (1873) 

 Bui'EONIN.iR ( 1873 ) . 

 Pandionest (1873). 

 Pernes§ (1873). 

 Elanill 1873). 

 Ictiniiell (1873). 

 Elani* (1873). 



Pycraft (1902). 

 Falconid.e. 

 Falconinse, part. 

 Polyborinje. 

 Falconinse, part. 

 Falconinse, part. 



BUTEORID,*. 



Pandioninse ? 

 Perninse, part. 

 Elaninse. 

 Milvinse, part? 

 Circinse,** part. 



* Changed in 1875 to Micrastures. 



t Changed 1875 to Herpetotheres. 



t My Pandiones originally included Elanoides, 

 which in 1875 I transferred to Pernes, where it 

 is placed by Mr. Pycraft. 



§ My Pernes included the genera Pernis, Baza, 

 Avicida ' Gymindis ' (Odontriorchis) , and Beger- 

 hinus. To these Pycraft adds, doubtfully, Pandion, 

 not being satisfied as to the propriety of separat- 

 ing it as a subfamily. Elanoides was added to the 

 group by me in 1875. 



II My Elani originally included only Elanus and 

 Gampsonyx, but Nauclerus was added in 1876. 

 Mr. Pycraft does not mention the last, but in- 

 cludes Machwt-haniphus, a genus which I had not 

 been able to examine. 



H This group includes lotinia and Earpagus, 

 the former being doubtfully referred by Pycraft 

 to the Milvinse (where most certainly it does not 

 belong) , while the latter is not mentioned by him. 

 Rostrhamus is also doubtfully referred by Pycraft 

 to the Milvinse, a group to which it seems to me 

 to be not at all nearly related. 



** Pycraft includes, besides Circus and ' Strigi- 

 ceps,' the genera UrotriorcMs and Geranospizias, 



