340 EMBEYOLOGY OF THE LOWEK VEETEBEATES CH. 



takes place fusion between adjacent centra so that compound centra 

 are produced. Most usually in this case it is the two centra (a, /?) 

 of one segment which undergo fusion but in some cases the posterior 

 centrum of one segment fuses with the anterior centrum of the next 

 so that the resulting compound vertebral centrum (/? a) belongs to 

 two successive segments. Again in some cases as exemplified by the 

 specimen figured (Fig. 166, A) three successive centra may undergo 

 fusion. 



Towards the front end of the tail and throughout the trunk 

 region the two centra of one segment undergo fusion but apparently 

 the hinder centrum has undergone reduction in size with the result 

 that its neural arch-element (2?) becomes displaced on to the top of 

 the smaller anterior element (A) (Fig. 166, B, C). 



Towards the extreme front end of the trunk the neural element 

 B becomes practically intervertebral in position (Fig. 166, D). It is 

 to be noticed that in these cases, where the element B has been dis- 

 placed, the bony splints which develop on its surface never spread 

 downwards from it, so that it remains throughout life without any 

 continuity of structure with the rest of the vertebra. 



Towards the tip of the tail there is apparently no regularity, all 

 kinds of fusions and modifications of the various elements taking 

 place. 



To sum up we see in Amia two complete potential vertebrae 

 corresponding to each segment, 1 each with its central, its neural and 

 its haemal elements. 



The two vertebrae of a segment may be represented by the 



A B 

 formulae a and /3 where a and /? are the centra, A and; 5-the neural 



~a b 



arch-elements and a and I the haemal arch-elements. In the trunk 

 region the ordinary compound vertebra may be represented thus 

 AB 



afl but in occasional cases fusions take place so as to produce vertel'rae 

 db 



BA ABA ARAB 



of the type /3a or of the type afia or a/Zap . Such a fusion as 



ba aba abab 



that shown in the last formula produces a \ertehra with a very lon^ 

 body upon which may persist four sets of neural and four sets of 



haemal element^. 



The great range 'of variation seen in Amia from the presumedly 

 original condition, even in different regions of the vertebral column 

 of one individual, emphasizes the need of much caution in the 

 laying down of ^<-in-r; t l principles n-^; inline; the composition of the 

 delinitive vcrtel.ra. It seems just iliahle to admit the two pairs of 

 neiir.il and two pairs of haemal arch elements into the general 

 1 Diplonpondyloni <-<>n<lition von Jhering. 



