88 E. T. NEWTON ON FISHES' TAILS. 



or 15), which gradually decrease in size as we trace them back- 

 wards. This tail is consequently truly heterocercal, although less 

 obviously so externally than in some other forms. It corresponds 

 to the period of the embryo flounder, stage 7, but with the 

 embryonic caudal fin obliterated. The Amia has the fleshy lobe 

 of the tail externally very nearly equal above and below the middle 

 line, but internally the ossified vertebral column is more suddenly 

 bent upwards than it is in Lepidosteus, and is continued by a 

 cartilaginous style into the uppermost border of the tail, so that 

 the latter is almost wholly below the vertebral column. The 

 haemal arches form such a regularly continuous series that one can- 

 not say definitely which should be regarded as hypural plates. 

 About 16 or 17 of the hsemal spines support the tail fin-rays, and 

 the terminal 8 have no corresponding neural arches. This tail 

 makes a much nearer approach to the homocercal than any we 

 have yet considered, the outer form showing but little inequality ; 

 but still there is a slight preponderance of the upper part of the 

 fleshy portion. It seems to me, therefore, to represent a somewhat 

 later stage of development than the Lepidosteus, but nevertheless to 

 be a truly heterocercal tail. Stage 10. 



6. Teleostei. — All the members of this group of the bony fishes 

 appear to have homocercal tails, and it is generally said to be 

 one of the characteristics of the Teleostei. But, as we have already 

 seen, there is considerable variation in the structure, although this 

 variation may be within the limits of the homocercal group. 



The tail of the eel seems to be the nearest approach to the 

 diphycercal form. 



The variation in the structure, as already pointed out, chiefly 

 concerns the number of hypural bones developed, there being 

 but one hypural in the cod, two in the stickleback, and seven in 

 the sprat. 



The homocercal tails also differ in the manner in which the up- 

 turned end of the vertebral column is ossified. The stickleback 

 has this portion of the notochord ossified in one piece, and in the 

 adult there is no part of the notochord unossified. In the sprat, 

 four or five bony segments are to be distinguished at one stage of 

 development, but the end of the notochord is never altogether 

 ossified. The salmon's tail has likewise several distinct vertebral 

 segments in the upturned portion, and, according to Bruch's figure 

 (loc. cit.), the unossified part of the notochord shows transverse 

 lines, indicative of segmentation. 



