of the Fishery Board for Scotland. 



263 



The heads of these ribs overlap the anterior dorsal, or rather external 

 surface of the basal pieces, and lie immediately behind the bony haema- 

 pophysial bar, which, though small in the more anterior, affords, from its 

 increasing size, a considerable support to the more posterior ribs. The 

 connection between these ribs and the basal pieces is firm, although not 

 intimate, and as seen by examination of the fry is secondary. 



Behind the 24th vertebra there are fourteen more ribs on each side, but 

 these are not attached to basal pieces (which disappear at this point), nor 

 directly to the haemapophyses. They ars simple rods of bone attached at 

 their proximal ends by ligament only to the hfemapophyses, and are there- 

 fore quite movable (PI. XV. fig. 1). 



The most anterior pair, the 25th (that of the 27th vertebra), are 

 attached to the haamapophyses close to their bases, but just as from that 

 point onwards to the end of the abdominal cavity the haemal spines elongate, 

 their tips (instead of the vertebral column) now forming the dorsal wall of 

 the abdominal cavity, so the ribs shorten; their attachment consequently 

 becomes gradually shifted down from the base of the haemal arch to near 

 the tip of the haemal spine (PI. XV. fig. 1). Their proximal ends run up for 

 a short distance behind and partially within the tips of the haemal spines. 

 While their position is not absolutely constant, it is generally found that 

 the last (36th) rib, which is considerably shorter than the 35th, does not 

 overlap the haemal spine, but arises ventral to it, and the heads of this 

 pair — and, to a certain extent, of two or three in front of it — lie close 

 together, and are attached to each other in the middle dorsal line of the 

 body cavity by the connective tissue. This last pair of ribs lies close behind 

 the vent. 



Transverse Processes. 



There appears to be connected to the anterior series of ribs a correspond- 

 ing series of elongated appendages (PI. XVI. figs. 4, 5, 9, 14, r.ap.). 



These are firmly attached to all the fixed ribs (except occasionally the 

 last pair). They originate just below the head of the rib, and pass directly 

 outwards and backwards, their tips lying within the lateral line. The 

 first two vertebrae, however, while not having ribs, have these appendages 

 (as transverse processes) directly articulated with the centrum. Where 

 the fixed ribs end the appendages again appear as direct vertebral trans- 

 verse processes, and quite unconnected with the free ribs; but here, while 

 arising at first from the side of the centrum, close to the base of the 

 haemal arch, their origin gradually moves to a more median position, fol- 

 lowing, in fact, the latero-ventral septum already described. 



Going backwards, they are found (gradually shortening in length) 

 present as far as the 38th vertebra (sometimes the 37th), holding a median 

 lateral position on it, but behind this point they no longer appear as ossi- 

 fications, being represented merely by a continuously shortening and 

 thickened thread of the connective tissue. 



An examination of herring fry makes more clear the condition of these 

 rib appendages. In the young fish they more closely approach the con- 

 dition of transverse processes of the vertebrae, holding, though already 

 attached to the ribs, a more distal position from them. The rib-head in the 

 youngest herrings examined by me, was found in some cases to lie along- 

 side of, but not actually in contact with, the basal piece, the rib developed 

 apart from the basal piece, and apparently not having yet at this stage 

 become attached to it. The head of the rib is recognised by its appear- 

 ance as a slight thickening, but it does not actually at this stage com- 

 mence there, for there is a prolongation of the head as a narrow rod, which 

 2l 



