vii EVOLUTIONARY ORIGIN OF LIMBS 447 



have retained its paired fins in a relatively primitive and inefficient 

 condition. 



Finally there are great physiological difficulties in the way of 

 accepting the lateral-fold hypothesis. There are no more fundamental 

 characteristics of the Vertebrate body than the arrangement of its 

 longitudinal muscles in segmental masses along each side of the body, 

 and the position of its skeletal axis, its central nervous system and 

 its main arterial trunk in the region of the mesial plane. It is quite 

 clear that such a creature is built for swimming by waves of lateral 

 flexure after the manner of an Amphioxus, a Lamprey or a Lung-fish. 

 Any new swimming organ that became evolved in primitive Verte- 

 brates must have had some advantage over, or at least not interfered 

 with, this primitive method of swimming. It is difficult to see how 

 the supposedly ancestral lateral fold could possibly have complied with 

 these conditions. The suggestion that the lateral fold may have 

 functioned at first, as a balancing organ or as a " bilge keel " will 

 not bear examination from the point of view of elementary physics. 

 Rabl suggests that the two lateral folds may have acted primitively 

 as a kind of parachute and that they became muscularized at their 

 anterior and posterior ends, the intermediate portion undergoing 

 atrophy (thus originating the two pairs of limbs). The skeletal 

 elements on this view would also develop at the ends of the ridge 

 first, and spread backwards (pectoral fin) or forwards (pelvic fin). 

 Thus would be explained the reversal of the position of the anterior 

 and posterior edges of the two fins in e.g. C'eratodus. Such an 

 explanation however fails entirely to meet the difficulty that there 

 exists not merely an antero-posterior reversal in the structure of the 

 two fins but also a dorsi-ventral one. 



The Gill-septum Hypothesis. — This hypothesis was based by 

 Gegenbaur (1872) on facts of adult anatomy. In some of the 

 Elasmobranehs (Pristis) the central gill ray attached to the branchial 

 arch is enlarged and the rays next to it have come to have their 

 bases of attachment shifted secondarily from the arch on to this 

 enlarged ray, so as to produce an arrangement recalling the biserial 

 archipterygium of Ceratodus with its central axis and lateral rays ; 

 Gegenbaur suggests that the archipterygium with its limb girdle 

 has in fact been evolved out of such an arrangement of rays 

 attached to a branchial arch and that the limb itself is serially 

 homologous with the gill septum. 



In support of this view it is pointed out that branchial arch and 

 limb girdle are each in early stages of development in the form of a 

 continuous curved rod of cartilage ; that this becomes usually 

 segmented in the case of the branchial arch but that even in the 

 girdle it also shows traces of segmentation in some ancient fossil 

 forms (Pleuracanthids, Acanthodians) ; that in some cases the peri- 

 chondrium of the pectoral girdle is known to be innervated by that 

 typical branchial nerve the Vagus ; that in the lower forms the 

 trapezius,one of the muscles associated with the fore-limb ; is innervated 



