OF GALLINACEOUS BIRDS AND TINAMOUS. 175 



of its circumference : all the upper portion is formed by the frontal, which is somewhat 

 bevelled, especially in H. varius. I am not certain whether there was any portion of 

 the nasal gland occupying a superorbital position. The general form of the skull in 



this smaller Hemipodius {H ?) (PI. XXXIV.) is much more like that of the common 



Quail than that oi H. varius (PI. XXXV.), which is more Pigeon-like and, indeed, has 

 its occipital plane much more basal than it is in the small kind. The whole aspect of 

 the skull and face of the Hemipodii is, to my eye, very much like that of the nearest 

 living relative of the Dodo, viz. Didunculus strigirostris. In both of these Hemipodii 

 the parietal region is evenly convex ; but whilst in the small kind the parietals and 

 frontals pass insensibly into one another in the adult bird, as in the Quail, they retain 

 their separate convexities in H. varius (PI. XXXV. figs. 3 & 4). This condition agrees 

 with what is found in the Pigeon, the Lapwing, and the Grouse (Pis. XXXVI. & 

 XXXVII.) . In H. varius the temporal fossa (PI. XXXV. fig. 4, sq.) is composed of 

 three parts, as in the Dodo ; indeed, considering the small size of the bird, these struc- 

 tures are singularly like what we see in that large extinct form. These parts are less 



distinct in the small Hemipodius ?, and are more in the condition of the temporal 



fossa of the Quail. In both kinds the postfrontal process of the alisphenoid and the 

 zygomatic process of the squamosal are very rudimentary. In the Quail and its large 

 congeners, the " Gallin2e " generally and the " CracidiB," they are long and splintery, 

 and in full age become fused together where they meet over the somewhat strong 

 temporal muscle. Here again the Hemipodius favours its columbine and pluvialine 

 relations (PI. XXXVII. figs. 4 & 9, sq.). The degree in which the lateral and basal 

 elements of the occipital sclerotome assist in the formation of the single condyle is very 

 different in the various families of birds. In the highest ornithic type, the Crow 

 (Corvus), the basioccipital takes the greatest share, the condyle being an evenly convex 

 oval, its long diameter lying in the an tero -posterior direction. In the " Gallinse " the 

 condyle is nearly bifid, being much more indebted to the lateral elements than that of 

 the Crow, and having its largest diameter transverse. In the Pigeons (PI. XXXVII. 

 fig. 6, o.c.) this takes place in a less degree; whilst in the " Struthionidse " and in 

 Syrrhaptes the condyle is only obscurely dimpled (PI. XXXVI. figs. 1 &3). The shape 

 of the condyle is predetermined in the primordial cartilaginous skull, as may be seen in 

 embryo birds before ossification takes place, and whilst the notochord lies in its axis. 



In the Lapwing (Vanellus cristatus) the single occipital condyle (PI. XXXVII. 

 fig. 1, o.c.) is a very neat hemisphere ; whilst in the Hemipodius (Pis. XXXIV. & XXX\ .) 

 its shape is almost entirely like that of the Crow. Here we are strongly reminded of 

 that marvellous little isomorph of the small Gallinacese, the Petroicabicolor (see Osteol. 

 Catal. Mus. Coll. Surg. vol. i. p. 299, no. 1584). Certainly this latter bird is the most 

 aberrant of the soft-billed songsters ; but whether it can claim any more than a general 

 ornithic relationship to the Hemipodius it is impossible to say. We shall see, as we 

 proceed, that \he. vomer oi Hemipodius is the exact counterpart of the same bone in the 



VOL. V. PART III. ^ * 



