2G4 



Appendices to Fifth Annual Report 



meets with the head of the appendage; this appearing as a rounded 

 thicker termination of the delicate stem (PI. XVI. fig. 14). Going back- 

 wards, we find this short rod or bar becoming thinner, and about the 25th 

 vertebra its central portion disappears, the appendage now arising from the 

 vertebra, and the rib showing merely a small pointed remnant of the con- 

 necting bar projecting from its thickened head (figs. 14a, 14&). 



The distance of the appendage from the rib (although connected with it), 

 as well as its complete independence of it on the first two vertebrae, and 

 the latter series of ribs, seems to point to its connection with the rib being 

 a secondary one. 



It has been stated that these processes project directly outwards, their 

 tips terminating internal to the position of the lateral line. If the stratum 

 of red muscles in this region be stripped off, a series of over-lapping, tri- 

 angular-shaped pieces of connective tissue are seen lying along the side 

 of the fish. These are found to be partly connective tissue, with more or 

 less of cartilaginous element deposited in it, and supported by a pair of 

 short (in the herring about 6 mm. long) cartilaginous rods widely sepa- 

 rated at their posterior (distal) extremities, which form the base of the 

 triangle, but conjoined towards their anterior ends (internal in position), 

 which forms the apex. The slightly cartilaginous connective tissue filling 

 up the space between the two arms forming this angle overlaps the apex of 

 tli 3 succeeding one. On the internal surface of the apex is a short groove, 

 and into this fits the fine terminal tip of the lateral appendage of the rib 

 (or transverse process). The connective tissue, in fact, surrounding 

 the tip of the appendage is undergoing the ossification to which in the 

 herring there seems throughout such a strong tendency. 



The Neural Arches. 



The whole of the vertebras bear neural spines. The anterior neural arch 

 and spine are stronger than the succeeding eighteen or twenty, and the 

 length is also less, the neural spines becoming longer and longer as far as 

 about the 12th vertebra. From thence they retain a similar length as far 

 the 30th, after which they shorten somewhat, and like the haemal spines 

 come to lie with a more pronounced backward slope. 



Up to about the 30th the neurapophyses of each vertebra are separated 

 to their tips from each other, beneath by the spinal cord, above by the 

 ligamentous cord and connective tissue; behind this point, however, both 

 sides of the spine become completely fused together, and the spinal cord 

 lies in a dorsally closed canal. To the 24th, or even 28th vertebra the 

 neural arches may be removed complete from the cup-shaped cavities on 

 the centrum, with which they articulate, although after the 24th this 

 cavity rapidly becomes shallower and less circular in outline, but beyond 

 this point the arches lose their articulation, and become perfectly fused 

 to the vertebrae. 



Each neurapophysis of the 1st vertebra shows a very slight iawardly 

 and backwardly directed process above and partly roofing the neural canal, 

 while their diverging bases [this term being employed for the part laterally 

 enclosing the neural cord, and lying immediately above the rounded nodular 

 articular 1 head '] are gradually and slightly expanded in an anterolateral 

 direction. Close to their articulation (which, owing to the shortness of 

 the vertebra, may partially overlap the basioccipital) is given off an out- 

 wardly and backwardly extending long fine appendage running between the 

 myotomes, but at a deeper level than will be found to be the case further 

 back (PL XVI. figs. 4, 5, 10-12, n.ap.). The second pair of neurapophyses 



