I 



PHYLOGENY. 99 



made with the ages successive steps of progress from 

 lower to higher conditions. This progress has not been 

 without exception, since, as regards the construction 

 of the scapular arch, the Mammalia have retrograded 

 from the reptilian standard as a whole. 



In subsequent pages I shall take up the lines of the 

 classes separately. 



b. The Line of tJie Pisces. 



The fishes form various series and subseries, and 

 the tracing of all of them is not yet practicable, owing 

 to the deficiency in our knowledge of the earliest or 

 ancestral forms. Thus the origins of the three sub- 

 classes, Holocephali, Dipnoi, and Elasmobranchii, are 



I lost in the obscurity of the early Paleozoic ages. The 

 genus Paleospondylus of Traquair from the Carbonife- 

 rous probably represents an Agnathous type from 

 which all fishes may have sprung, although the genus, 

 as now known, has not sufficient antiquity to claim 

 this place. It may be a near descendant of the amphi- 

 oxus. 



A comparison of the four subclasses of fishes shows 

 that they are related in pairs. The Holocephali and 

 Dipnoi have no distinct suspensory segment for the 

 lower jaw, while the Elasmobranchii and Teleostomata 

 have such a separate element. The latter, therefore, 



, present one step in the direction of complication be- 



'I 5'ond the former. It is, however, asserted by Huxley^ 

 that the absence of suspensorium is due to its appro- 



|| priation by the hyoid arch in the Holocephali, and its 

 rudimental condition in the Dipnoi. If this be the 

 case, the Holocephali and Dipnoi are peculiar speciali- 



^ Proceedings Zoological Society, London, 1S76, p. 45. 



