360 PROCEEDINGS OF THE ACADEMY OF [1887. 



radii or digits, in the transition from the many-rayed fishes. 

 This is a very significant fact and is of striking importance, as 

 indicating that there is a certain general correspondence between 

 the number of nerve pairs and the number of digits in the hepta- 

 dactyle, (^=what was formerly considered the pentadactyle) limb of 

 higher vertebrates, since the identification of extra, but vestigiary 

 radial, ulnar, tibial and fibular digits by Bardeleben. The seven 

 digits of the manus never much exceed the usual five, to six pairs 

 of the proplexus, while the seven digits of the pes do not much ex- 

 ceed or fall below the six to nine nerve pairs of the most usual type 

 of postplexus. We saw too, that in those cases where an excessive 

 number of digits were developed in the fore-limb, as in the case of 

 Raia, there was an exact correspondence in the number of nerve 

 pairs of the proplexus. The exactitude of this correspondence is in 

 fact, apparently, in proportion to the degree to which the digital 

 elements radii (of the pro- meso- and metapterygium) have retained 

 their archaic composition, relation, want of torsion, etc., in either 

 limb, 



3. Fusion of radii to form the j^^o- meso- and metapterygium and 

 their inexact homology. 



On the basis of the doctrines established by Dohrn, through onto- 

 genetic research, it is quite safe to assume with him, that the three 

 basal elements of the limb in Elasmobranchs have been derived from 

 the primitively separate cartilaginous radii, developed in the meso- 

 blastic tissue between the muscular limb-buds thrown oflf by the 

 somites. The different genera of Selachians, however, show that 

 the pro- meso- and metapterygium, as suspected by Wiedersheim, 

 are probably not exactly homologous, because the mesopterygium is 

 not always present, and when present, upon comparing any pair of 

 genera, it will be found that in no two do the number of radii present 

 in the pro- meso- and metapterygium correspond. This difference 

 is apparently due to the fact that the number of radii in the whole 

 limb, in different genera, is not constant, as already pointed out. 

 Furthermore, it is evident that the pro- meso- and metapterygium 

 respectively, cannot be of the same morphological value in different 

 genera, if the same number of somites do not take a share in the 

 formation of each of these three parts in different genera. It follows 

 from this that neither the uniserial nor biserial archipterygium of 

 Gegenbaur and Huxley can be made to yield such a fixed hypothet- 

 ical type as will lead up to the various modifications of the paired 



