EEPORT ON THE SPHENISCID^. 15 



longi colli muscles, while in the succeeding vertebrae they form large, laterally compressed 

 plates of bone of larger size, which present no trace of bifidity. These processes 

 in Eudyptes chrysocome have their bases attached to the whole length of the vertebrae 

 to which they belong. They diminish slightly in size from the fourth to the eighth 

 dorsal vertebra. 



The following variations with respect to the bodies of the dorsal vertebrae occur 

 in different species. 



In Ajytenodytes, Pygosceles, and Eudyptes chrysolophiis the body of the first dorsal 

 vertebra is destitute of the articular surfaces for the reception of the heads of the first 

 pair of ribs, these bones in the species named articulating only with the transverse 

 processes of the first dorsal vertebra. 



The hypapophyscs in all three varieties of Eudyptes chrysocome, as also in Apteno- 

 dytes are attached by their bases to the whole length of the bodies of the vertebras to 

 which they belong, whUe in Pygosceles, Eudyptes clirysolopthus, and every species of 

 Spheniscus the bases of the hypapophysial spines are attached only to the anterior half 

 of their respective vertebrae. In Pygosceles and Eudyptes these processes are relatively 

 larger than in other species. In Sp>hemscus demersus, Spheniscus magellanicus, Spheniscus 

 minor, and Aptenodytes the h}'papophysis of the fifth dorsal vertebra differs from the 

 corresponding process of other species in being bifid. In Sj^heniscus minor the last dorsal 

 vertebra possesses a small hypapophysial spine. 



The transverse processes are long, pointed, and flattened in the dorsal region. With 

 the exception of those belonging to the first dorsal vertebra they are all directed slightly back- 

 wards. The transverse processes of the first dorsal vertebra difi"er from those of the others, 

 inasmuch as they more nearly resemble those of the last cervical than those of the other 

 dorsal vertebrae. They are, moreover, inclined slightly forwards instead of backwards. The 

 articular surfaces for the tubercles of the ribs are situated on the free extremities of the trans- 

 verse processes, except in the case of the first dorsal vertebra, in which they are situated mid- 

 way between the base and apex of its transverse process. The apex of the transverse process 

 of the last dorsal fuses with that of the first lumbo-sacral vertebra. With this exception 

 all the transverse processes are quite distinct and in no way connected with one another. 



The articular p>rocesses present the usual arrangement ; the posterior articular pro- 

 cesses are destitute of the bony spines which characterise certain of them in the 

 cervical region. 



The spines of the dorsal vertebrae are broad, flattened, quadrilateral plates of bone, and 

 thus difier much in form from those of the cervical region, with the exception of the last, 

 the spinous process of which closely resembles that of a dorsal vertebra. The spine of 

 the last dorsal vertebra like the transverse process is anchylosed mth that of the first 

 lumbo-sacral, and forms a portion of the bony ridge formed by the fusion of the spines of 

 another, and secondly, the the lumbo-sacral vertebrae. 



Comparative Remarks. — The peculiarities of the dorsal portion of the vertebral 



