ON THE THEORY OF THE VERTEBRATE SKULL 595 



"' But it is necessary to distinguish carefully between what we call vertebras in the 

 adult fish, that is to say, those osseous or cartilaginous pieces intended for the 

 support of the whole body, and more particularly of the spinal cord, and such 

 vertebral divisions as we find in embryos. These last are the general fact, the 

 expression of a constant law according to which all the Vertebrata are developed. 

 The vertebras of adult fishes, on the other hand, are solid rings, whose presence 

 ■depends on the type peculiar to each species ; consequently their form, and the 

 substance of which they are composed, vary in almost every species. 



" The vertebral divisions ' appear very early in the Coregonus — almost at the 

 same time as the notochord ; and when the dorsal groove begins to close, they are 

 fine lines, caused, as it would appear, by a greater accumulation of embryonic cells, 

 which, hke transverse septa, traverse the entire mass as far as the notochord. 

 These divisions extend forwards, as far as the neighbourhood of the auditory 

 vesicles, but there never exists the smallest trace of them in the head itself At 

 first they are visible only in the middle of the body ; by degrees they move 

 forwards, as far as close to the ear, and backwards, towards the tail, as far as it is 

 formed ; but they invade its extremity only when it has attained its full length 

 relatively to the body. At first these lines are all straight and perpendicular to the 

 axis of the chorda ; but by degrees, and in proportion as development advances, 

 they become oblique and bend, forming an angle whose apex is directed forwards, 

 and corresponds exactly to the median line of the notochord." 



They eventually become the intermuscular septa. Vogt goes on to say, — " The 

 typical structure of the Vertebrata, then, consists solely in these rings of separation, 

 •which are formed around a notochord, and nowise in the development of a distinct 

 head, or of other solid pieces of the skeleton, such as osseous or cartilaginous 

 vertebrffi," illustrating his case by the Amphioxus. 



Each osseous vertebra corresponds to two vertebral segments, namely, to the 

 half of that which precedes, and to the half of that which follows a metasomatome ; 

 for it is where the latter reaches the notochord, that the centra and arches take 

 their origin. The centra arise as a double ring of cartilage, internal and external 

 to the sheath of the notochord. The intervertebral spaces always correspond with 

 the middle of the interval between two intermuscular septa, each of which is con- 

 .sequently inserted into the middle of a centrum, while the superior and inferior 

 .arches are developed in their plane. They are ossified only long after the centra, 

 which arise as bony rings around the notochord. The intervertebral ligaments are 

 formed from the sheath of the notochord. 



4. Prof Owen (' Principes d'Osteologie Comparee,' 1855, p. 184) affirms that " In 

 ■osseous fishes the centrum is ordinarily ossified from six points, of which four 

 begin in the bases of the two neurapophyses and of the two parapophyses, but the 

 terminal concave plates of the centrum are ossified separately." 



It is not stated on what fish the observations on which this latter assertion is 

 based were made. Prof Williamson has already shown ('On the Development 

 of the Scales and Bones of Fishes,' Phil. Trans. 185 1) that it is inconsistent with 

 the structure of the adult vertebra ; it is not supported by any of those writers who 

 have directly observed the development of the vertebrse of Teleostean fishes, and 

 it is negatived by the observations of Vogt just cited and by my own. 



Development of the Spinal Column of Batrachia. 



I. A nura {Dughs, ' Recherches sur les Batraciens,' 1835). — In the first period 

 the notochord appears to be divided transversely into " rondelles " or vertebra, but 



^ Metasoraatomes, their interspaces being the somatomes. 



Q Q 2 



