744 



anus, i. e. at the base of the tail where the effects of caudal flexion are 

 most potent, to form the caudal vein (fig. 5 Cheiropterygia). 



This re])lacement may theoretically either result 1) from the actual 

 fusion of the posterior cardinals in the median line, or 2) from the per- 

 sistence of one of them in the median line, or 3) from their replacement 

 by a third and therefore new median channel. Actually, it is probable 

 that processes 1) and 3) occur in all cases of ontogeny, process 1) 

 occurring medianly and 3) anteriorly; however, according to Ho di- 

 stette r, the post-caval of birds entirely results from process 3), and 

 the post-caval of mammals, according to Gegenbaur, from processes 

 2) posteriorly and 3j anteriorly. Wliether the method of replacement 

 of the posterior cardinals by the ])ost-caval which occurs in ontogeny 

 truly represents the phylogenetic process (which was jjrobably that indi- 

 cated above for ontogeny — 1) medianly and 3) anteriorly) is an ojien 

 question, but however this may be, it can hardly be doubted I think but 

 that the primitive cause of the replacement was that which I have in- 

 dicated and which is necessarily correlated with the evolution of loco- 

 motor limbs. 



Referring back to the case of the dorsal position of the posterior 

 cardinals of fishes, the question as to why they are dorsally rather than 

 ventrally situated in the median line is not a difficult one. As in the 

 problem just considered, so here the answer is that the blood pursnes 

 this particular course because it is the path of least resistance. In 

 the case of the posterior cardinals, the blood returned from the region 

 of the tail passes forwards under and in close apposition to the verte- 

 bral column, in fact, in the haemal arch of each vertebra. Now in fishes 

 the posterior extremity of the coelomic cavity is situated ventrally i. e. 

 in the vicinity of the anus, and the dorsal side of the coelom gradually 

 rises from this posterior ventral limit towards the dorsally-placed verte- 

 bral column; hence, owing to this conformation the dorsally-placed 

 caudal vein must inevitably bifurcate into two vessels which continue to 

 lie dorsal to the body-cavity, or, in other words, follow the course of the 

 nearly straight vertebral column. 



In vertebrates above fishes the tail region is often considerably re- 

 duced, but still in the vast majority of cases the channels returning the 

 venous blood will continue to lie dorsally to the body-cavity, since this 

 dorsal course is, as in fishes, more in line with the caudal vein than 

 the ventral, and all animals, save the anurous Amphibia (which have a 

 piscine larva however), possess tails of sufficient size to render the flow of 

 blood in the caudal vein an important factor in determining this direction 

 of flow. In mammals too, the upwardly-inclined iliac veins would tend to 

 effect a junction with a dorsally-placed vein rather than with a sub-ab- 



