1 907.] TO THE OSTEOLOGY OF BIRDS. 373 



Families Phytotomidse, Pitticlse, and Tyrannida?. This 1 have 

 done because there seems good reason to regard these three 

 nearly related Families as close allies of the Diacromyodean 

 Passeres. 



As FLirbringer — improving on the earlier work of Huxley (2), 

 Sclater (7), and ISTewton (3) — has alread}'' suggested, the Passeri- 

 formes should, I think, be regarded as forming four Suborders : — 



1 Eurylsemi. 



2 Oligomyodi. 



3 Ti-acheophones. 



4 Diacromyodi = Oscines. 



With the Eurylsemi must be included the Eurylsemidse, 

 Cotingidse, Pipridee, and PhUepittidee : the whole forming one of 

 four great branches of a common stem. The other three give 

 rise, respectively, to the Tracheophonse, the Tyraunine forms 

 (including the Phytotomidse and Pittidfe), and the Diacromyodean 

 Passeres. 



The Phytotomidfe, I would remark, are cei^tainly more nearly 

 related to the Tyrannidaj than to the Pittidse ; and this fact is 

 largely borne out by the structui-e of the skull. 



Osteologically, the Oligomyodean and Diacromyodean suborders 

 have much in common. The skeletons of the Tyrannidae and of 

 the Muscicapidte, for instance, present very close resemblances. 



The formation of the Suborders just referred to is based largely 

 on myological characters, such as are afforded by the syrinx and 

 the wing-muscles. All that concerns the former is so well known 

 that it will not be necessary to repeat it here. As to the wing- 

 muscles, however, a few words are necessary. 



In earlier papers (5) I have already shown that the deltoideus 

 tnajor (pars longa and pai-s h7^evis) affords useful characters for 

 systematic purposes ; and the condition of these muscles in the 

 groups now under discussion offers further confirmatory evidence. 

 Thus, pvit briefly, in all save the Diacromyodean Passeres the 

 pars hrevis fails to reach the ectepicondyloid process of the humerus. 

 In the Eurylfemid group it is very short, being confined to the 

 upper end of the humerus ; and this is true also of the Pittidfe 

 in the Oligomyodean group ; but in the Tyrannidae this muscle 

 extends more than halfway down the shaft. In this the Tyran- 

 nidpe agree with the Tracheophone Passeres, from which however 

 they differ in the syringeal muscles, osteology, and pterylosis. 

 Nevertheless, the differences between the wing-muscles in the 

 Tyrannidae and Pittidae are differences of degree and not of kind. 



Diacromyodean Passeres alone seem to have retained the del- 

 toideus hrevis in its entirety : but even here it shows, in some 

 Families, a very marked tendency to decrease, while, in others, 

 on the other hand, it is relatively of great size. In the Musci- 

 capidse and Hirundinida? it would seem to have presei-ved its 

 most primitive character. In the former, this muscle is very 

 thick and fleshy throughout its whole length up to the point of 



