which can only be considered a decided retrograde from, instead of an 

 improvement upon, the better classifications which preceded it. In the 

 first place, the genus Pandion is made to form a "suborder" — Pandiones! 

 This would not be so bad, all things considered, were it not that th© 

 genus Polioaetus is included in this so-called "suborder", the latter 

 being a very near relative, probably a subgenus, of Haliaetus, one of 

 the group Buteones'iof the subfamily Buteoninw, and not at all allied to 

 Pandion (group Pandiones of Buteoninw), though it represents that genus 

 in its group ; that is, is analogous, but not affined to it. In the next 

 place, Serpentarius {— family Serptntariidw) and Cariama {— family 

 Cariamidce) are placed in the " subfamily " Polyhorinw! The subfami- 

 lies which Mr. Sharpe recognizes are the following: (1) Polyborinm, (2) 

 AccipitrincB, (3) Buteonince, (4) Aquilince, and (5) Falconinw. 



In this connection, it may be well to call attention to the importance 

 of distinguishing between evidences of affinity and those of mere anal- 

 ogy in the birds of this family. Certain types o f teleological modifi- 

 cations are repeated in members of different subfamilies, and in different 

 groups of one subfamily, to such a degree of perfection, that, if we were 

 to follow external appearance only, we would not hesita te to place 

 them near together in a systematic arrangement. This is what has 

 led to such confusion and such utterly unsatisfactory results as have 

 characterized most attempts at a natural classification. Among the 

 more prominent instances ©f analogy between members of different 

 groups or subfamilies, or even different families, the following may be 

 mentioned : The family Gathartidce is reproduced in the vultures of the 

 Old World (vulturine series, group Buteones, subfamily Buteoninw, fam- 

 ily Faloonidce) ; the Gypogeranidce and Cariamidce, in a less degree, by 

 Polyboroides and Qeranospizias {Buteones); the Strigidw, in a very slight 

 degree, by Circus {Buteones, Buteonince), Micrastur {Micrastures, Fal- 

 conincB), and Pandion {Pandiones, Buteonince).* The Falcones are rep- 

 resented in the Buteoninw by Ictinia and Sarpagus {Buteones); also 

 by Bam {Pernes) ; one of the Polybori {Ibycter) by Bostrhamus {Buteones) ; 

 the Micrastures aie repeated in the genera Circus and Nisusf {Buteones), 

 and JSerpetotheres very nearly imitated by Circaetus {Buteones). Taking 

 the Buteonince alone, the Pandiones are mimicked by certain Saliaeti 

 {Buteones) ; Certain Pernes {Flanoid es) by Milvus mdlSfauclerus {Buteones) ; 

 and others of the same group {Bam and Aviceda) by Harpagus and 

 Ictinia. 



In three widely-distinct " series " of genera in the group Buteones, we 

 find a very peculiar type of modification, viz, the excessive abbreviation 

 of the outer toe. This occurs only in Eeteropus (an " aquiline " form , 

 with densely-feathered tarsus), Polyboroides (a long-1 egged terrestrial 

 form, with reticulated tarsus), and Geranorpozias (similar to the last, but 

 with scutelate tarsus) ; these latter two differ from the first in being 

 of terrestrial habits, and in a very remarkable teleological modification 

 of the libio-tarsal joint, whereby it can be bent with ease in both direc- 



The following tabular arrangement expresses the equivalents of the 

 two subfamilies and several groups, adopted in this memoir in the 

 numerous " subfamilies," or " families " of leading authorities, and suc- 



inner 

 age 



the outer toe reversible and in lacking after-shafts to tbe feathers. i, , ■ i 



t In general form, Micraslnr exactly repeats the genus Nisus, and has, besides, the tacial 

 rnff of the genus Circus. 



