STERXUM. 463 



form to;»ether with the haemal arches a ch)sed hsenuil canal which superficially 

 resembles that in Elasmobranchii. 



In Lepidosteus and Amia, especially the former, the same phenomenon 

 is still more marked ; and in Lepidostens it is easy, in passing backwards, 

 to trace the ribs bending ventral-wards, and uniting ventrally in the caudal 

 region to form, with the htemal processes, a complete hajmal canal. 



It might have been anticipated that the Teleostean Ganoids would 

 resemble the Teleostei, but, from an examination of adult Teleostei, it 

 would seem to be clear that the relations of the ])arts are the same as in 

 Elasmobranchii, i.e. that the ribs have no share in forming the haemal 

 canal in the tail. Aug. Miiller and Gdtte have however brought embryo- 

 logical evidence (though not of a conclusive character), to shew that in the 

 embryo the ribs really fuse with the haemal processes iu the tail, and so 

 assist, as in the Ganoids, in forming the haemal canal. Giitte moreov^er holds 

 that the ribs in Elasmobranchii are not homologous with those of Teleostei 

 and Ganoids ; but that the haemal arches in the tail are homologous in the 

 three groups. 



Without necessarily following Gotte in these views it is worth pointing 

 out that the undoubtedly close atfiuity between the bony Ganoids and the 

 Teleostei is in favour of the view on the haemal arches of Teleostei at 

 which he has arrived on embryological grounds. 



In Amphibia the formatit)n of the ribs from the connective tissue of the 

 intermuscular septa, their secondary attachment to the transverse processes 

 of the neural arches, and their subsequent separation was first clearly 

 established by Fick (No. 431), whose statements have since been confirmed 

 by Hasse, Born, &c., and in part by Giitte, who holds however that, though 

 converted into cartilage independently of the transverse processes, they 

 are formed in membrane as outgrowths of these processes. 



In the Amniota the ribs are also independently established (Hasse and 

 Born), though they subsequently become united to the transverse processes 

 and to the bodies of the vei'tebrae, or to the ti*ansverse processes only. 

 This junction is however stated by the majority of authorities, never to 

 be effected by the fusion of the cartilage of the two parts, but always by 

 fibrous tissue ; though Hoftman (No. 435) takes a different view on this 

 subject, holding that the ribs are at first continuous with the intervertebral 

 regions of the primitive cartilaginous tube surrounding the notochord. 



Sternum. In dealing with the development of the sternum it 

 will be convenient to leave out of consideration the interclavicle or 

 episternum which is, properly speaking, only part of the shoulder- 

 girdle and to confine my statements to the sternum proper. 



This structure is found in all the Amniota except the Ophidia, 

 Chelonia, and some of the Amphisbaena3. 



From the older researches of Rathke, and from the newer ones of 

 Gotte, etc., it appears that the sternum is always formed from the 

 fusion of the ventral extremities of a certain number of ribs. The 

 extremities of the ribs unite with each other from before backwards, 

 and thus give rise to two cartilaginous bands. These bands become 

 segmented off from the ribs with which they are at first continuous, 

 and subsequently fuse in the median ventral line to form an unpaired 

 sternum. The Mammalian presternum (manubrium sterni) and 



