30 n.AIKISHKS (HETKROSOMATA) 



In uioo Kyle (njooii) published ;i valuable paper on the classification of the 

 Hetcrosomata, making use of several new characters of some importance He 

 arranged the various groups as follows : 



1-aiiiilics. SubfcUiiilifS. 



I Hippoglossinas, 



,„ . , Pleuronectina;. 



1 leuronectula' „ , u u 



H ippoglosso-rhombmae. 



Khombina?. 

 I Solein,T. 

 Soleid:c Achirina?. 



( Cynoglossinae. 

 Solei-pleuronectinDP. 

 Incerts sedis ? 

 In uiio Regan (lyioB) proposed an entirely new classification of the group, based 

 on a careful study of the anatomy, and especially of the osteology, of a number of 

 genera. He divided the order mto two suborders, I^settodoidea and Pleuronectoidea, 

 the former containmg a single family, Psettodida?. The Pleuronectoidea were further 

 divided into two main divisions, Pleuronectiformes and Sola-iformes, corresponding 

 to the Pleuronectidae and Soleidae of Jordan and Evermann. each of w-hich contained 

 a dextral and a sinistral family. Each of the families of Pleuronectiformes was split 

 up mto three subfamilies. Expressed in tabular form, Regan's arrangement of the 

 Heterosomata stands as follows : 



Order Heterosomata. 

 Suborder Psettodoidea. 

 Family Psettodidae. 

 Suborder Pleuronectoidea. 

 Division Pleuronectiformes. 

 Family Bothidse. 



Subfamily Paralichthinae. 

 Platophrinje. 

 BothinK. 

 Family Pleuronectidae. 

 Subfamily Pleuronectina?. 

 ,, Samarinae. 



Rhombosoleinae. 

 Division Solaeiformes. 

 Family Soleid^. 



Cynoglossida;. 

 In a later classification Regan (1929) omitted the suborders and divisions, recog- 

 nising five familes : P-settodida", Bothid^, Pleuronectid;c, Soleidae and Cynoglossid^e. 

 The subfamilies of Bothidae and Pleuronectidae recognised in his original classification 

 were retained, but the South -African genus Paralichthodes was removed from the 

 subfamily Samarinae and placed in a separate subfamily, Paralichthodin;e.' In his 

 most recent classification of fishes Jordan (1923) has followed Regan in the main as 

 far as the FTatfishes are concerned, merely raising most of the subfamilies to the rank 

 of families. Apart from the creation of another subfamily in the dextral Pleuro- 

 nectidx', the I^a-cilopsettin;p, my own work on the group- has pro\'ided no valid 

 reason for interfering with Kegan's latest arrangement, which would appear to be a 

 perfectly natural one. The reasons for adopting this classification, many of which 

 .ippear in Regan's earlier paper, may now be briefly outlined. 



' Sec also Regan (1920, p. ::i3J. 



= This has been mainly on the " I'loundtrs ". whirh form the snbji-rt of this volume. The 

 Sok-ida' and Cynoglossid.x- have not yet been txtensivcly studied. 



