PHTLOGENT OF THE PAL.EOGNATH^ AND NEOGNATHJE. 223 



In the sketch the supra-coracoid foramen is not indicated, neither is there any 

 indication of clavicles. 



In Casuarius the coracoid is relatively much shorter than in Dromceus. Its breadth 

 equals its length. The whole sternal border rests within the coracoid groove. Unlike 

 Dromceus, the pro-coracoid is large and there is a small pro-coracoid fenestra. In the 

 young Casuarius the pro-coracoid is ligamentous. The supra-coracoid foramen is large 

 and receives numerous pneumatic apertures. The scapula resembles that of Dromceus 

 in the general form and in the size of the acromion. 



In Rhea the coracoid is long and comparatively slender, tapering rapidly from its 

 sternal border forwards to the supra-coracoid foramen. The pro-coracoid is ligamentous, 

 leaving a conspicuous notch on the mesial border of the macerated coracoid. There is 

 generally a supra-coracoid foramen, but this never receives pneumatic apertures. The 

 whole sternal border rests in the coracoid grooves. At the base of the coracoid and on 

 its dorsal surface is a large pneumatic aperture. This does not seem to be represented 

 in any other Palceognathce except, and feebly, some Crypturi. 



The scapula is peculiar, curving from the coracoid first inwards then backwards. 

 The glenoid surface for the humerus lies at the distal end of the coracoid instead of at 

 its anterolateral angle. 



In Strutliio the size of the pre-coracoid is enormous, extending backwards as a long 

 finger-like process ultimately to reach and fuse with the internal angle of the base of 

 the coracoid, enclosing in this way a huge fenestra. The sternal (epi-coracoid) border 

 of the coracoid is of great width. The supra-coracoid foramen is wanting. We might 

 mention here that the pre-coracoid of S. molybdophanes appears to be peculiar in its 

 greater relative slenderness. 



In Apteryx the pre-coracoid seems to be wanting only in A. oiveni, and in this species, 

 it would seem, at all stages of its growth. In the remaining species it is represented in 

 the adult by ligament only ; so that in the macerated skeleton the pre-coracoid fenestra 

 is represented only by the deep notch on the mesial border of the coracoid. In a 

 skeleton of A. australis mantelli this notch has been nearly filled up by a secondary 

 extension of bone in the coracoid of the left side. The supra-coracoid fenestra is 

 minute or absent in all but A. oweni, in which it is large. This point, remembering 

 the absence of the pro-coracoid in the latter and its presence in the remaining species, 

 is interesting. 



In the Dinornitliiclce. the coracoid has reached a state of extreme reduction, being 

 rod-like and merging almost insensibly with the scapula. 



In the JEpyornithiclce, in the form of the pro-coracoid, it is intermediate between 

 Rhea and Dromceus. 



The coracoid of the Crypturi differs from that of its congeners — the flightless Palceo- 

 gnathce — chiefly in that it has a well-developed acro-coracoid. There is a well-marked 



2 i2 



