1887.] NATURAL SCIENCES OF PHILADELPHIA. 351 



grow out close behind the pectoral limbs, with little or no interven. 

 ing space between them, but notwithstanding this fact, and in spite 

 of an extensive subsequent translocation of the pelvic limbs forward 

 into a position in advance of the base of the pectorals, the paired 

 nerves which go to the pelvic limbs retain their primary origin be- 

 hind those which pass to the pectorals. In the adult physoclists, 

 therefore, the nerves going to the pelvic limbs, cross below those 

 going to the pectorals, on their way to the pelvic limbs. This re- 

 tention of the original nerve origins is in itself the best proof that 

 we can depend upon to give us a clew to the groups of somites from 

 which a given limb has arisen. 



In many cases the origins of the paired nerves passing from the 

 cord are much further forward than the foramina or intervertebral 

 intervals which give them exit. This difficulty is probably quite 

 explained away by the manner in which the vertebral canal grows 

 in length compared with the cord. It is found, in fact, that the 

 vertebral canal grows in length much faster, in many forms, than 

 the cord, after a certain period. This causes the origins of the spinal 

 nerves from the cord to appear as if they had been drawn forward 

 some distance in advance of their points of exit from the sides of 

 the vertebral canal. That this is a true explanation is proved by 

 the fact that Kolliker has found the cord extending the entire length 

 of the vertebral canal in the human embryo of three months, while 

 the writer has found the same condition in the embryos of Cetaceans, 

 two inches in length. It is, therefore, obvious that the cauda equina 

 in these cases is developed at a later period, and as a result of the 

 gro^vth in length of the spinal or vertebral canal at a more rajiid 

 rate than that of the included cord itself. Similar phenomena oc- 

 cur in the cases of certain fishes (Lophius) and Goette has described 

 the process in Bomhinator igneus. In this last case, however, there 

 is more or less positive atrophy of the posterior end of the cord in 

 the course of development, so that only about 14- pairs of spinal 

 nerves can be finally identified. In Mola^ not only the cord, but 

 the tail is also aborted to such an extent that only a very short, almost 

 occipital, cord remains, the paired nerves passing directly to the 

 lateral musculature of the vertical and paired fins, after forming a 

 dorsal cauda equina. The cord in the long, flagelliform, reduced 

 tail of the two Lyomeri, viz : Gastrostomus and Ophiognathus, shows 

 unmistakable signs of atrophy or degeneration, in that the cord in 

 the tail becomes so reduced to a mere flattened filament, that it is 



