No. 3.] THE VERTEBRAL COLUMN OF FISHES. 139 



the growth of the chorda. But the arcualia are bound to- 

 gether by the bone which I have described, and it is this bone 

 too which leads to the constriction of the chorda. All this 

 happens before a spicule of other bone is visible. 



It might be supposed that the bone which has been de- 

 scribed by Gadow and Abbott and that described by myself 

 are identical. I have carefully considered this possibility, but 

 I cannot reconcile their views and my own. 



The renewed study of my preparations in the attempt to 

 test the statements of the authors referred to has led to the 

 recognition of some things to which I had attached perhaps too 

 little importance, and which I at one time thought might offer 

 an explanation of the views of those writers. There do exist 

 what appear to be outgrowths from the arcualia, and these out- 

 growths extend from one arcuale to the other, connecting all of 

 those of each disc ; but these outgrowths lie against the elastica, 

 betivccn it and the first formed bone. In the anterior dorsal 

 region of a specimen 44 mm. long there is observed such a 

 layer of cartilage, although it may be seen in younger speci- 

 mens. It is thickest near the articular ends of the centra, 

 while in the intermediate regions it becomes very thin, a single 

 layer of cells. Immediately between the bases of the various 

 arcualia of each vertebral centrum or disc it seems to be wholly 

 wanting, thus producing gaps in what forms practically a 

 sheath of cartilage around the notochord. This reminds us of 

 the sheath of cartilage which surrounds the notochord of the 

 young Lepidosteus, except that in Amia the cartilage is not 

 continuous from one vertebra to those adjoining. 



As to the origin of the cartilage, it undoubtedly develops 

 after the layer of bone has been deposited and spreads from 

 the bases of the arcualia. When the bone is first laid down 

 over the elastica it either lies close against this sheath or there 

 are to be seen between the elastica and the bone, a very few 

 nuclei which closely resemble those of the bone-cells already 

 enclosed. A little later the cells lying under the bone increase 

 in number; and gradually they take on the appearance of carti- 

 lage cells surrounded by their matrix. 



It seems to me that this thin layer of cartilage represents 



