BIRDS OF NORTH AND MIDDLE AMERICA. 171 



decidedly so; equal to or slightly exceeding in length the combined 

 length of basal and middle phalanges of middle toe. 



The absence of obvious rictal bristles is the only external character 

 that I am able to discover which will serve to distinguish the Icteridse, 

 as a group, from the Fringillidfe. It is true that none of the Icterida^ 

 have the bill notched (i. c., the maxillary tomium notched subtermi- 

 nally), but neither do many genera of Fringillidse. With the latter 

 group the Icteridse agree in the following characters : 



(1) The strongly deflexed or angulated rictal portion of the com- 

 missure (not always found in Fringillidae, howe\er). 



(2) The abortion of the first primary. 



(3) The complete, or nearlj^ complete, separation of the basal pha- 

 lanx of the middle toe from that of the inner toe (being united to the 

 outer toe l)y about half the length of the basal phalanx, as in Corvidse 

 and other groups). 



(■±) In having the planta tarsi closelj' joined along their anterior 

 edge to the acrotarsium, and with only the extreme lower portion 

 divided. 



In both groups the inner plate of the planta tarsi is depressed con- 

 spicuously below the level of the acrotarsiuni, the posterior margin of 

 which stands out as a prominent I'idge (more conspicuously so than in 

 most CorvidaB), and those of opposite sides are in close contact along 

 their posterior margin, forming a sharp posterior edge to the tarsus. 

 In both groups the proportionate length of the toes themseh-es and the 

 relative development and curvature of the claws varies considerably 

 and to about the same extent. 



From the Sturnidfe and the Ploceidse, which are undoubtedlj', next 

 to the Fringillidte, the most nearlj^ allied groups, the Icteridte may^ be 

 distinguished by the complete abortion of the first primary, both of 

 those groups being "ten-primaried." 



From the Corvidw, the Icteridse may be distinguished by more 

 numerous characters, among which may be mentioned (1) the posses- 

 sion of only nine obvious primaries; (2) the absence of obvious rictal 

 bristles, and (3) the undivided lower portion of the planta tarsi. 



Although so nearl}^ allied to the Fringillidai that only a single exter- 

 nal character seems available for its diagnosis, the Icteridse neverthe- 

 less constitute a well-circumscribed group, there being not a single 

 genus whose proper reference to it can be seriously questioned. At the 

 same time it is a group presenting most extreme types of teleological 

 development or adaptive modification, the strictlj- arboreal oi'opendolas 

 and caciques (genera Oymnostino2)s, Zarhynchus, Clypicterus, Cacicus^ 

 etc.) representing one extreme and the terrestrial meadowlarks 

 (Sturnella, Trupialis, and Leistes) the other, peculiar specializations 

 being manifested in the boat-tailed Qu/scali and spiny-tailed DoU- 

 clwnyx. With its limited representation in North America it would 



san^i;; 



