THE THEORY OF THE VERTEBRATE SKULL. 293 



since, in the Lampreys, the arches of the vertebrae arise inde- 

 pendently within the lamella?, which the investing mass of the 

 notochord has sent out to embrace the central parts of the 

 nervous system. Still less weight can be attached to the cir- 

 cumstance that not unusually, even when both sphenoidal centra 

 are present, only one pair of the corresponding ala? appears ; 

 while, in other cases, two pair of aire and only one central part 

 are present, since the caudal vertebra? of Mammalia usually 

 exhibit no traces of arches, and the Lampreys have such arches 

 without centra. On the other hand, the circumstance is impor- 

 tant that the basi-sphenoid, although it arises within the invest- 

 ing mass of the notochord, is not developed around this (as, so 

 far as our present observations go, even the most posterior 

 caudal vertebra? are), but in front of it, in a process of the 

 investing mass ; and that the body of the presphenoid is no 

 longer developed, even in a part of this mass (except in a few 

 Mammalia), but arises quite independently of it. Hence, the 

 two sphenoids no longer agree perfectly with vertebra? in their 

 development — the anterior diverging more widely from the ver- 

 tebral type than the posterior. 



" (9.) Yet the two sphenoids, like the proper vertebra?, still 

 embrace segments of the nervous tube (such as is formed by the 

 brain and spinal marrow, at any rate in the early stages of 

 development), and they constitute, as the vertebra? at first nor- 

 mally do, open rings, or rather segments of rings, round that 

 tube. The ethmoid, however, at no time surrounds a segment 

 of the nervous tube in question ; but, in a few animals only, im- 

 perfectly includes, by its hinder part, two anterior prolongations 

 of that tube, w T hence the olfactory nerves arise. Its mode of de- 

 velopment, and its ultimate form likewise, are of such a character 

 that it no longer offers any special resemblance to a typically- 

 formed vertebra. Nevertheless, considering that it arises from 

 *a part of the prolonged investing mass of the notochord — viz., 

 from the anterior, early-coalescing parts of the two trdbeculse — 

 and that its body (the pars perjoendicularis) presents even a cer- 

 tain resemblance to the last caudal vertebra? of many Birds and 

 osseous Fishes, it may well be considered to be a modified vertebra. 

 We may look at it, in short, as the representative of only the 



