THE SHAPES OF VERTEBRATE SKELETONS 207 



tmch bones as those called sesamoid; together with ethers 

 too numerous to name. 



Again, in the course of evolution, both as displayed in the 

 Vertebrata generally and in each vertebrate embryo, three 

 skeletons succeed one another the membranous, the car* 

 tilaginous, and the osseous. These substitutions take place 

 variously and unsystematically. While one part of a skele 

 ton retains the membranous character, another part of the 

 same skeleton has become cartilaginous. At the same time 

 that certain components have become partially or completely 

 ossified, other components continue cartilaginous or mem 

 branous. Further, though there is a general succession of 

 these stages, the succession is not regularly maintained ; for 

 in many cases bones are formed by the deposit of osseous 

 matter in portions of the membranous skeleton, which thus 

 do not pass through the cartilaginous stage. &quot; Nor,&quot; says 

 Prof. Huxley, &quot; does any one of these states ever completely 

 obliterate its predecessor ; more or less cartilage and mem 

 brane entering into the composition of the most completely 

 ossified skull, and more or less membrane being discoverable 

 in the most completely chondrified skull.&quot; And then, too, 

 the processes of chondrification and ossification often proceed 

 with but little respect for the pre-existing divisions ; but 

 severally may result in the establishment of two parts wher&amp;lt;j 

 there was before one, or one where there were before two. 

 Now wholly incongruous as these facts are with the hypothe 

 sis of an archetypal skeleton, they are quite congruous with 

 the mechanical hypothesis. This shows us why, in the 

 course of evolution, a feebly- resisting membranous structure 

 came to be replaced by a more-resisting cartilaginous struc 

 ture, and this, again, by a still-more-resisting osseous struc 

 ture ; and why, therefore, these successive stages succeed one 

 another, as it seems so superfluously, in the vertebrate em 

 bryo. And it further shows us why there is irregularity in 

 the succession ; seeing that the varying mechanical ac 

 tions to which the varying habits of the Vertebrata have 

 4G 



