48G y\r. C. T. Regan on (lie Origin and 



upper eye ; but it seems wrong to speak of any part of the 

 latter bone as a new formation, least of all that part which 

 has the same position and the same relations (except to the 

 eye) as it would have if tlie skull were symmetrical. 



For a long time the flat-fishes were regarded as asym- 

 metrical Gadoids, but in the latter the absence of spinous fin- . 

 rays, the large*nuinber of rays in the pelvic fins, and the-'^^'*^"^ 

 indirect attachment of the pelvic bones to the cleithramay be '"'' 

 regarded as primitive features, and it is probable that these 

 fishes have evolved from generalized Iniomi, such as the 

 Aulopidse. In tlie Heterosomata, on the contrary, spinous 

 fin-rays are present in Psettodes, the least specialized member 

 of the order, the pelvic fins are never more than 6-rayed, and 

 the pelvic bones are directly attached to the cleithra. Pset- 

 todes is simply an asymmetrical Percoid ; about the first ten 

 dorsal rays are spinous, the caudal has 17 rays, 15 of which 

 are branched, and each pelvic fin is formed of a spine and 5 

 soft rays. The moutii, the skull (except for its asymmetry), 

 the pectoral arch, and the vertebral column are all quite 

 Percoid. 



In other Pleuronectoids all the fin-rays are articulated, and 

 in many of them the pelvic fins are 6-rayed, with tlie anterior 

 ray simple. I am unable to confirm Boulenger's statement 

 that an additional ray is present in Hippoglossus, and it is 

 clear to me that the anterior pelvic ray of this and other 

 genera with 6-rayed pelvic fins corresponds to the spine of 

 Psettodes, and that the formation of joints in response to 

 mechanical requirements has reconverted spines into articu- 

 lated rays in the dorsal and pelvic fins of the Heterosomata, 

 as in the case of the epaxial rays of the homocercal caudal 

 fin*. 



Thilo t and Boulenger \ consider that the Zeidae are nearly 

 related to the Heterosomata ; I cannot find any anatomical 

 evidence in support of this idea, I much more readily sub- 

 sciibe to Boulengei's view that tHe Upper Eocene Amphistium 

 is allied to the symmetrical ancestor of the flat-fishes, for in 

 my opinion this fish is a Percoid, which should probably be 

 placed in the family Scorpididse near the existing Psettus, 

 or may perhaps be related to Platax. Thanks to the courtesy 

 of Dr. Smith Woodward, 1 iiave been able to examine the 

 two examples of Amphistium paradoxum in the British 

 Museum. The caudal fin has 17 principal rays, 15 of which 

 are branched, in addition to a few graduated rays above and 



* Eegan, Ann. & Mag. Nat. Hist. (8) t. 1910, p. 357. 



t Zool. Anz. 1902. pp. 305-320. 



t Ann. & Mag. Nat. Hist. (7) x. 1902, pp. 295-304. 



