1890.] THE STRUCTURE OF PSOPHIA. 339 



position among Schizognathous birds ; and I maintain that the struc- 

 ture of the skull is at least not opposed to such a conclusion. 



In Psophiu the omentum is very much leduced and the gizzard is 

 attached to the j'^vietes b}' an almost vertically running sheet of 

 membrane ; the elongated sternum completely covers the lobes of 

 the liver, and the gizzard itself only just reaches beyond it. In 

 these particulars Psophia especially resembles Cariama and Chunyu 

 and the Rallidse, and differs in the most pronounced fashion from the 

 Gruidse. I have not yet dissected many Limicol^e from this point 

 of view ; but in Hcematopus ostralef/ns the disposition of the 

 omentum is more like that of the Cranes, though it is more reduced 

 than in that grouj). 



The above comi)arison of Psophia with other forms shows that it 

 cannot be closely united with any other genus of those referred to. 

 It ap])ears to me to be most widely removed from the Liuiicolse, 

 though it is connected with this family by CEdicnemus, which should, 

 in my 0])inion, be removed from the Limicolse. 



The Limicolse may, from their skull-characters, be thus defined :— 



Schizorhinal birds with the maxillary process of the nasal directed 

 forwards. Sujiraorbital margin with a sharp edge marked above by 

 considerable furrows for the nasal glands. Foramen magnum on the 

 under surface of the skull ; occipital foramina ])resent. Vomer 

 truncated or ending in a concave margin. Interorbital septum 

 largely unossified. Lachrymal and ])refrontal process of ethmoid 

 fused to form a com])lete ring of bone. Maxillo-])alatines very small 

 and fused with ])aiatines, invisibly or nearly so from beneath. Post- 

 orbital angle aud ])ostfrontal process distinct ; temporal fossa com- 

 mencing behind former process and visible on the occipital surface of 

 skull. Basipterygoid ])rocesses usually present. 



The family Rallidae shows the following characters : — 



Holorhinal birds with the maxillary process of the nasal facing 

 outwards. Supraorbital margin rounded and without depressions 

 for nasal glands. Foramen magnum on the jjosterior face of the 

 skull ; no occipital foramina. Vomer pointed in front. Palatines 

 narrow in front, wide behind ; maxillo-palatines large and swollen, 

 quite conspicuous from below. Lachrymals quite free from pre- 

 frontal process of ethmoid. Interorbital septum incomj)letely ossified. 

 Articulation of quadrate covered by a descending ])rocess of squa- 

 mosal. No postorbital angle ; temporal fossae just reach the occipital 

 face of skull. No basipterygoid processes. 



If the members of these two famihes were the only " Gralline " 

 birds known, it is obvious that' there would be no difficulty in 

 accepting Prof. Huxley's ' arrangement of them into two sections — 

 Charadriomorphas and Geranomcrphse ; but a consideration of other 

 forms, including Psophia, appears to me to render this arrangement 

 impossible. 



The Cranes themselves are the first stumbling-block. They 

 agree with the Limicolse in many, perhaps most, characters, but in 

 others they agree with the Rallidse. 



1 " On the Classificatiou of Bii-ds, &c.," P. Z. S. 1867, j). 457. 



