aintatomt of the shoe-bill. 695 



projecting anterior end of the keel of the sternum. There is no 

 trace of a median process opposite the hypocleideum between the 

 arms of the furcula. Anchylosis with the keel may be regarded 

 as an adaptive character. It is as complete in the Pelican as in 

 Balceniceps : it exists in some of the large Storks, but in Storks 

 and Herons genei-ally the connection is by ligament. In Scojncs 

 the furcula is far short of reaching the keel. In 6'cojnis, Storks 

 and Herons including Cancroma, the furcula is moi-e "U "-shaped ; 

 the median forward process between the arms of the " U" is 

 absent in Sco2>n,s and Storks as in Balcenicejys, but is well-marked 

 in Herons. 



The proximal end of each clavicle where it i-eaches the coracoid 

 has a very strong flange (text-fig. 132, A, Cl.a.) which articulates 



Text-fig. 132. 

 :-Ac. 



Shoulder-girdle Articulation in Bcdaniceps and Scopiis. 



A. Salcenicejis. B. Scop^ts. 



Ac. Acrocoracoid process of Coracoid. 



Cor. Coracoid. 



Pc. Procoracoid process of Coracoid. 



CI. Clavicle. 



CI. a. Acrocoracoid process of Clavicle. 



Sc. Scapula. 



directly with the anterior border of the coracoid. A similar 

 flange is present in Scopus (text-fig. 132, B, Cl.a.), and in its place 

 there is a minute flat articular facet in Tantalus, but it is absent 

 in Storks generally and in Herons including Cancroma, the 

 clavicle being attached to the acrocoracoid only by strong liga- 

 ments. The development of this flange in Balcenice2JS is a very 

 strongly marked chai'acter, but too much weight canirot be laid 

 on this similarity with Scojnis, for the acrocoracoid flange of the 

 clavicle is equally well marked in the Pelican, in Plotus, in. 

 Cormorants and Gannets, and in Birds-of-Prey. 



