344 



Part III. — Eighteenth Annaul Report 



ribs. The secondary ribs arise from the transverse processes or the 

 neural arches. Beginning from the second abdominal vertebra — some- 

 times they appear to be present on the first, they arise from the neural 

 arches, then their points of origin pass on to the centrum, until on the 

 seventh or eighth vertebra they arise from transverse processes which 

 are placed more to the ventral than dorsal surface of the vertebra. In 

 Citharus, the secondary ribs are continued on to the caudal vertebrae. 

 This disposition of the vertebra? was remarked by Costa (8) in Rhomboid- 

 ichthys, and he classified them into three — cervical, dorsal, and caudal, 

 by analogy with the terms employed for higher vertebrates. These 

 terms are obviously misleading, because we should then have ribs on the 

 cervical vertebra?, and it is better to call the vertebra 1 which form the 

 dorsal wall of the abdominal cavity abdominal, and to subdivide them 

 when necessary into those with ha?mapophyses and those without. 



According to Gunther's definition (21), the caudal region would begin 

 with the first vertebra which bears the ha?mapophyses. According to 

 this the abdominal cavity in the above species would be limited to the 

 first four to six vertebrae from the skull, whereas it is in reality pro- 

 longed much further back. A better sign of the delimitation of the 

 abdominal cavity and of the commencement of the caudal vertebra? in 

 the above species, as in the groups previously mentioned, is the first 

 interhaemal spine. This is large, and forms the posterior boundary of 

 the abdominal cavity. 



Ainoglossus and Rhomboidichthys form a good example of the value of 

 this character for classification. Formerly these were separated from 

 one another, and Arnoglossus was considered to be more closely related 

 to Lepidorhombus (13). More recently (50) the latter has been placed 

 along with the Topknots, and the two former brought close together 

 under one genus (Platophrys), in spite of the great external differences 

 between them. As mentioned above, Arnoglossus and Rhomboidichthys 

 have a similar arrangement of the vertebra?. Lepidorhombus, on the 

 other hand, is like the Turbot and Topknots in this character, and is 

 therefore widely removed from the two former. In another important 

 character, viz. the ventral fins, Arnoglossus and Rhomboidichthys are 

 also connected and different from the other genera of the Turbot-group, 

 so that their characters are very clear and definite. 



The arrangement of the vertebrae which is found exceptionally in the 

 Turbot-group becomes the rule in the Sole-group. Both in the Cyno- 

 glossinos and the true Soles, all the species examined showed the hcema- 

 pophyses in the abdominal vertebra 3 . These Jwrnapophyses are rela- 

 tively much shorter than those of the caudal vertebra?, and in Solea (Fig. 

 9) there is a gradual transition from the one to the other. 



In addition to the differences in the formation of the abdominal 

 vertebra? in the various groups, there are also important differences in 

 other characters connected with the abdominal cavity. The first inter - 

 haemal spine which forms its posterior boundary is large and strong in 

 the Plaice- and Halibut-groups, extending in a groove along the first 

 spine almost to the centrum of the first caudal vertebra. The outer 

 end of this spine is very sharp, and may even project through the skin. 

 This character — the presence of an "anal spine" — is employed by Giiuther 

 and the American writers as a specific character, but the appearance of 

 the spine externally may depend greatly on the condition of the speci- 

 men. Thus Gunther declares that there is no anal spine in the Long 

 Rough Dab (D. platessoides), whereas the American writers say there is. 

 The latter again declare that P. cynoglossus has an anal spine, whilst 

 P. microcephalus lacks it, and both they and Gunther state that there 

 is no anal spine in the Halibut. As a matter of fact the first inter- 



