CLASSIFICATION 41 



wanting, the rays being inserted directly on the hypercoracoid. In the Samarinae the 

 hypocoracoids are expanded, but in all the remaining subfamilies these bones are 

 narrowed forward below (Fig. 25). 



In the vertebral column, the parapophyses of the praecaudal vertebrae are separate 

 in the Pleuronectinae, Poccilopsettina" and Rhombosoleinae, the last pair are connected 

 by a bridge in the Paralichthodinae, and in the Samarinae all the parapophyses are 

 united to form closed hsmal arches, which bear the slender ribs at their extremities. 



In his classification of the Flatfishes, Kyle (iqoob) made use of one very interesting 

 character which has been already mentioned, namely, the arrangement of the olfactory 

 laminae in the nasal organs. The differences in the structure of these organs in various 

 Flatfishes had been previously noticed by Bateson (1889), and this character was used 

 by Regan (1910B) to differentiate the dextral Pleuronectidae from the sinistral Bothids. 

 Kyle stated that in Psettodes and the Bothida?, as well as in the Soleidae and Cyno- 

 glossidae, the laminae radiate from or are arranged transversely to a central rachis. 

 This rosette-like form of nasal organ is characteristic of the majority of .symmetrical 

 Bony Fishes.^ In the dextral Pleuronectidae, on the other hand, he found that in all 

 the genera examined the laminae were only slightly raised above the level of the basal 

 supporting membrane, and were arranged parallel to each other and to the main axis 



Fig. 36. — Nasal organ of ocular side of a, Hippoglossus hippogtossus ; b, Scophthalmus niaximits. 

 [After Kyle,] «., nasal bones ; n.s., nasal sacs ; ol., olfactory lamina;. 



of the body, without a central rachis (Fig. 26). I have carefully tested this character 

 in a number of genera and species not examined by Kyle, and find that, although it 

 is of considerable value in defining genera or even subfamilies, it is of less importance 

 than was formerly supposed. The form and arrangement of the laminae are described 

 in the diagnoses of the genera concerned, but some indication of the variation 

 encountered within the different families and subfamilies may be given here (Fig. 27). 

 In many of the genera of Bothidae the lamina; radiate from or are arranged transversely 

 to a central rachis of some length. In others (e.g. Thysanopsetta, Hippoglossina) 

 the laminae have a radiating arrangement, but the rachis is very short. In others, 

 again (e.g. Tcrniopselta, Botltiis, Mancopsetta, Phrynorhombiis) the central rachis has 

 disappeared, and the laminae, although well-developed, may be much reduced in 

 number, and either arranged in slightly radiating form or parallel to each other and to 

 the main axis of the body. 



Among the Pleuronectida; the parallel arrangement described by Kyle is by no 

 means constant. In the subfamily Pleuronectinae (in which the laminae are often well 

 raised) it occurs in all the genera with the sole exception of Atheresthes, in both species 

 of which the two series of laminae are arranged transversely to a central rachis.'^ 

 Jordan and Evermann (i8g8, p. 2609) regarded this genus as one of the most primitive 



^ For a good account of the nasal organs in Bony Fishes generally, see Bume (1909), Derscheid, 



(1024). 



- 1 am indebted to Prof. C. L. Hubbs for drawing my attention to this fact. 



