522 



exhibiting the typical structure : remarking that " mammals, 

 and reptiles show the rule of connexion, and fishes the exception." 

 Thus in the case of the scapular arch, the evidence afforded by 

 lishes is held of great weight, because of their archetypal regularity ; 

 while in the case of the pelvic arch, their evidence is rejected as 

 exceptional. But now, having, as he considers, shown that these 

 bony frames to which the limbs are articulated are modified ha3inal 

 arches, Professor Owen points out that the haemal arches habitually 

 bear certain " diverging appendages ;" and he aims to show that 

 the " diverging appendages " of the scapular and pelvic arches re- 

 spectively, are developed into the fore and hind limbs. There are 

 several indirect ways in which we may test the probability of this 

 conclusion. If these diverging appendages are " rudimental limbs '"' 

 " future possible or potential arms, legs, wings, or feet," we may 

 fairly expect them always to bear to the hajmal arches a relation 

 such as the limbs do. But they by no means do this. " As the 

 vertebra? approach the tail, these appendages are often transferred 

 gradually from the pleurapophysis to the parapophysis, or even to 

 the centrum and neural arch." (Arch, and Horn., p. 93.) Again, 

 it might naturally be assumed that in the lowest vertebrate forms, 

 where the limbs are but little developed, they would most clearly 

 display then- alliance with the appendages, or " rudimental limbs," 

 by the similarity of their attachments. Instead of this, however, 

 Professor Owen's drawings show that whereas the appendages are 

 habitually attached to the pleurapophyses, the limbs, in their earliest 

 and lowest phase, alike in fishes and in the Lepidosiren, are articu- 

 lated to the haemapophyses. Most anomalous of all, however, is 

 the process of development. When we speak of one thing as being 

 developed out of another, we imply that the parts next to the germ 

 are the first to appear, and the most constant. In the evolution of 

 a tree out of a seed, there come at the outset the stem and the 

 radicle ; afterwards the branches and divergent roots ; and still 

 later the branchlets and rootlets ; the remotest parts being the latest 

 and most inconstant. If, then, a limb is developed out of a " di- 

 verging appendage " of the ha3mal arch, the earliest and most con- 

 stant bones should be the humerus and femur ; next in order of 

 time and constancy should come the coupled bones based on these ; 

 while the terminal groups of bones should be the last to make their 

 appearance, and the most liable to be absent. Yet, as Professor 

 Owen himself shows, the' actual mode of development is the very re- 

 verse of this. At p. 16 of the Archetype and Ilomoloyies, he says: 



" The earlier stages in the development of all locomotive extremities are 

 permanently retained or represented in the paired tins of fishes. First tha 

 essential part of the member, the hand or foot, appears : then the fore-arm 

 u- leg, both much shortened, flattened, and expanded, as in all fins and all 

 embryonic rudiments of limbs : finally come the humeral and femoral seg- 

 ments ; but this stage I have not found attained in any fish." 



