22 NATURAL SCIENCE. Jan., 



else could epiphyses ever be discriminated ? Again, the circumstance 

 of a bone or cartilage making its appearance as a single element may 

 in any case be due to the junction of its incipiently distinct parts at a 

 period anterior to possible observation ; it may be connate, i.e., never 

 distinct to observation, though judged from analogy to be essentially 

 multiple. Of such rationally-inferred but invisible distinctness, we 

 have examples in botany. The results of development are surely as 

 much to be considered as are its earlier stages. 



As to the vertebral or non-vertebral nature of the skull, it seems 

 to us that such vertebral nature may be affirmed in one sense and 

 denied in another, according to the line of thought which is followed. 



It is undeniable that it is within the lamince dorsales of the embryo 

 that the cartilages and bones of both the brain-case and of the dorsal 

 portion of the spinal cord are alike formed. Certainly also the bones 

 of the skull — especially in the higher animals — present a singular 

 reminiscence of vertebra in the three serially successive arches they 

 form ; and if the essence of vertebrae consists in their being a series 

 of bony rings fitted together, and serially enclosing the nervous 

 centres, then, in that sense, the skull is in part composed of three 

 vertebrae. In some fishes {e.g., the sturgeon) the transition from the 

 spinal column to the skull is so gradual that the two are difficult to 

 distinguish. In the Silurian fish Bagrus (as Owen pointed out) the 

 vertebrae next the head are greatly expanded, and join each other by 

 suture, and this shows how undoubted vertebrae may simulate osseous 

 cranial walls. 



It is true that, unlike the nascent spinal column, the primitive 

 skull presents no serial segmentation, and that the cartilaginous 

 precursors of the cranial bones may even be divided medianly and 

 antero-posteriorly. Nevertheless, in the young axolotl, before the 

 base of the skull has become cartilaginous, an index of transverse 

 segmentation is to be traced in the soft tissue of that region, in spite 

 of the subsequent continuous chondrification of the base of the skull. 



But most striking of all is the return made^ of late years by Sir 

 Richard Owen's main opponent to the conception that serial segmen- 

 tation, however latent and disguised, extended primitively and funda- 

 mentally to quite the anterior end of the head, and that, judging 

 from Ampiiioxtis, the skull of the higher vertebrates may be repre- 

 sented as made up of something less than twenty serial segments, 

 each of which is equivalent to a spinal vertebra with its annexed parts. 



The conception of cranial vertebrae, then, like the conceptions of 

 serial, lateral, and general homology, are subjective apprehensions of 

 relations having an objective existence in nature. They are like 

 conception " types," which are not, of course, real objective 

 entities, as types, though they are conceptions which have none the 

 less an objective basis. The acceptance of the theory of evolution is 

 no bar to the reception of the view which represents all organic forms 



Proc. Roy. Soc, no. 157, p. 127. 



