1902.] OSTEOLOGY OP THE PALCONIFORMES. 313 



the Falconiform palate must be our justification foi- piu"suing this 

 mattei- a little further. 



The palate of the Oathai'tfe is undoubtedly of an extremely 

 specialized type, but traces of a like modification are not wanting 

 in a dii-ection where hitherto they have not been looked for — to 

 wit, in the Accipiti-es pi-oper. 



The palate of Elanus, as we have already described, is schizo- 

 gnathous, a fact first pointed out by Shufeldt ; but the schizo- 

 gnathism is of a specialized character, being due to the extreme 

 reduction of the maxillo-palatine processes. If the nasal septum 

 in the sliuU of Elanas be examined, it will first of all be remarked 

 that it is moi-e complete than in the Cathartfe, and next that, near 

 its posterioi- infeiior angle, it gives off a pair of small horizontal 

 processes, I'esembling those of Cathai-tfe, but relatively smaller ; 

 these almost touch the maxillo-palatines. A little increase in 

 the size of these sjdui'S, and the reduction of the anteiioi' portion 

 of the septum would give us the Cathartine palate. Thus, then, 

 the palate of Elanus must be held to represent the high-water 

 mark of specialization in the dii'ection of schizognathism in the 

 Accipitres. 



It is probable that the palate of Circaetus — or rather the 

 maxillo-palatines — represents the intermediate type from which 

 the extremes of schizog- and desmognathism in the group have 

 been derived. Furthei'more, it may profitably be compai'ed with 

 the paleeognathine maxillo-palatine of, say, the Tinamous. It 

 must be remembered that the great feature of the palseognathine 

 palate is the enormous size of the maxillo-palatines. In the highly 

 specialized Tinatnus this is much reduced, and in general shape 

 is not unlike that of Circaetus, but lacks its vertical plate of 

 spongy tissue. It is impoi-tant that the intermediate character 

 of the palate ^of Clrccoehcs shotdd be recognized, othei'wise we 

 commit ourselves to the admission that the desmognathous 

 palate of forms like the Falconidfe or Aquila, ior example, were 

 developed by the resuscitation of an almost defunct organ. In 

 Circaetus, then, the maxillo-palatines are represented by a paii- of 

 vertical plates of spongy tissue neai'ly meeting in the mid-ventral 

 line. They run backwards nearly as far as the antorbital plate 

 (prefrontal) ; and forwai-ds, then inwaixls towards the tomium, 

 so as to leave a palatal vacuity exposing the nasal septum as in 

 Elanus. Reduction of this type of maxillo-palatine gives us 

 the highly specialized type of Elanus, its further increase the 

 type seen in the Falconidte, notably through llilvago to Poli/boi'tcs 

 and Ibycter (PI. XXXIII. figs. 3, 5), wherein the maxillo-palatines 

 have attained a relatively enormous size. In the Falconince we 

 meet with an exactly parallel series, passing through Harpa to 

 HerjJetotheres (PI. XXXIII. figs. 2-4). In the higher Falcons an 

 inci-ease in size of the anterior nasal chamber, eventually, in 

 Falco, developing into a much inflated ossified bviUa, has brought 

 abou.t a considerable reduction in the size of the maxillo-palatine 

 processes ; the fenestrated bullae, seen in such perfection in 

 Proc. Zool. Soc— 1902, Yol. I. No. XXI 21 



