of the Fishery Board for Scotland. 



343 



The bone in front of the clavicle to which the laminal cartilage of the 

 ventral fins in the Turbot- group is attached requires special mention, 

 as it seems to have a special importance in the Bat-fishes. Its form is 

 that of a two- pronged fork, the base of which rests on the clavicle, the 

 upper prong passing forward into the substance of the tongue, and the 

 lower shorter fork bending obliquely downward and forward. In the 

 Tnrbot-group this latter prong is directly connected with the ventral 

 fins, in the other groups mentioned above it has no such connection, but 

 a ligament passes from its end to the end of the clavicle and thence to 

 the base of the ventral fin. It would seem to have an important 

 physiological significance, more especially in the Turbot-group, where the 

 bone is larger than in the other groups. It acts as a lever, the clavicle 

 acting as fulcrum, so that any movement of the fins which affects the 

 lower prong will also affect the upper, and thence also the tongue and 

 branchiostegal apparatus, the size of the mouth being thus affected. It 

 is interesting to note further that in the specialised rhomboid-forms, 

 the Topknots, the ventral fins are relatively nearer the mouth than in the 

 other forms, and this is connected with the more anterior position of 

 the pubic bone (Fig. 20). The habits of these forms are also specialised, 

 for, as shown by Cunningham and others, these fish swim, or rather hang 

 suspended in a vertical position, amongst the seaweed or rocks, and ma.y 

 use the ventral fins to keep themselves in position. 



Vertebral Column and Alimentary Canal. 



The formation and arrangement of the vertebrae and vertebral pro- 

 cesses furnish important evidence of affinities within the various groups 

 of flat-fishes, much more so than in the roimd-fishes. The vertebra^ are 

 divided into Lwo classes — abdominal and caudal, the differences between 

 whi^h are generally well marked. The former possess certain processes 

 projecting laterally and ventrally on each side, and called par apophyses 

 by Giinther, to which primary and secondary ribs are a ttached. In the 

 caudal region the parapophyses have joined ventrally to form the 

 ha niapophyses, which resemble greatly the neurapophyses. According to 

 Giinther, the abdominal region ends and the caudal region begins with 

 the first vertebra bearing the hsemapophyses. 



This distinction between the vertebra? applies generally amongst the 

 round-fishes, and also in the least specialised of the flat-fishes, i.e. in the 

 Halibut- and Plaice-groups. The abdominal vertebra, in addition to the 

 neurapophyses, is provided with ventrally projecting parapophyses, to 

 which are attached the primary (cf. Fig. 11, p.r.) and secondary ribs 

 (s.r.). The secondary ribs are short, and arise from the anterior verte- 

 bra?. The primary ribs are long, and extend posteriorly and ventrally, 

 those of the posterior abdominal vertebra? reaching as far as the first 

 interha?mal spine. This spine lies along the anterior surface of the 

 first ha?mal process, and thus corresponds to the beginning of the caudal 

 vertebra?. These latter have the ordinary structure, and are usually 

 provided with more or less prominent transverse processes. 



A similar disposition of the vertebra? is found in the Turbot, but its 

 other characters show this form to be highly specialised, and it is 

 probable that this disposition has arisen secondarily, i.e. has reappeared 

 after passing through a certain transitional stage. The reason for this 

 statement is that the forms which stand between the Turbot and the 

 most primitive group of flat-fishes have a different arrangement of the 

 vertebra?. In these, Citharus, Amoglossus, and Bhomboidichthys, the last 

 three to six abdominal vertebra? bear short ha?mapophyses and trans- 

 verse processes (Fig. 11), which replace the parapophyses and primary 



