THE PECTORAL UIHDLE. 493 



ventral or infra-clavicle (inter-clavicle), which meets its fellow 

 below. 



In most Fishes the primitive dermal scutes have become sub- 

 dermal membrane bones, and the infra-clavicle is usually not distinct, 

 but the two clavicles form the most important part of the mem- 

 branous elements of the girdle. Additional membrane bones (post- 

 clavicles) are often present behind the main row. 



The development of these parts in Fishes has been but little 

 studied. 



In Scyllium, amongst the Elasrnobranchii, I find that each half of 

 the pectoral girdle develops as a vertical bar of cartilage at the front 

 border of the rudimentary fin, and externally to the 'muscle-plates. 



Before the tissue forming the pectoral girdle has acquired the cha- 

 racter of true cartilage, the bars of the two sides meet ventrally by a 

 differentiation in situ of the mesoblastic cells, so that, when the girdle 

 is converted into cartilage, it forms an undivided arc, girthing the 

 ventral side of the body. There is developed in continuity with the 

 posterior border of this arc on the level of the fin a horizontal bar 

 of cartilage, which is continued backwards along the insertion of the 

 fin, and, as will be shewn in the sequel, becomes the metapterygium 

 of the adult (figs. 344, bp and 348, itij)). With this bar the remain- 

 ing skeletal elements of the fin are also continuous. 



The foramina of the pectoral girdle are not in the first instance 

 formed by absorption, but by the non-development of the cartilage 

 in the region of pre-existing nerves and vessels. 



The development of these parts in Teleostei has been recently investigated 

 by 'Swirski (No. 472) who finds in the Pike (Ksox) that the cartilaginous 

 pectoral girdle is at first continuous with the skeleton of the tin. It forms 

 a rod with a dorsal scapular and ventral coracoid process. An independent 

 mass of cartilage gives rise to a prsecoracoid, which unites with the main 

 mass, forming a triradiate bar like that of Acipen^er or the Siluroids. The 

 coracoid process becomes in the course of development gradually reduced. 



'Swirski concludes that the so-called prsecoracoid bar is to some extent 

 a secondary element, and that the coracoid bar corresponds to the whole of 

 the ventral part of the girdle of Elasmobranchii, but his investigations do 

 not appear to me to be as complete as is desirable. 



Amphibia and Amniota. The pectoral girdle contains a more or 

 less constant series of elements throughout the Amphibia and Am- 

 niota; and the differences in structure between the shoulder girdle 

 of these groups and that of Fishes are so great that it is only possible 

 to make certain general statements respecting the homologies of the 

 parts in the two sets of types. 



The generally accepted, view, founded on the researches of Parker, 

 Huxley, and Gegenbaur, is to the effect that there is a primitively 

 cartilaginous coraco-scapular plate, homologous with that in Fishes, 

 and that the membrane bones in Fishes are represented by the 

 clavicle and inter-clavicle in the Sauropsida and Mammalia, which 

 are however usually admitted to be absent in Amphibia. These 



