210 Cockerell — Some Notes on Fish Scales. 



Scales large, minutely ctenoid above 1. 



Scales minute 2. 



1. Lateral areas with very fine radiating stri;e . Syacium papillosum (L. ). 



Albatross Sta. 2405. 

 Lateral areas without such stripe . . . Citharichthys arctifrons Goode. 



2. Basal radii so many and close as to produce the effect of a fine 



striation ; scales broad 3. 



Basal radii not thus numerous 4. 



3. Apical teeth of scales on upper side of fish projecting beyond 



margin Platophrys constellatus Jordan. 



Magdalena Bay (Albatross). 

 Apical teeth of scales on upper side submarginal; so that the 



scales seem cycloid Platophrys ocellatus Agassiz. 



Albatross Sta. 2414. 



4. Basal radii many, about 15-20 on lower, over 20 on upper side; 



teeth of ctenoid scales few (about 5 to 7) and very sharp . . . 



Pseud opleuronectes americanus Wall). 

 Basal radii fewer 5. 



5. Lateral areas, free from radii, extensive, their circuli dense . . 



Notosema dilectum Goode and Bean. 



Lateral areas, free from radii, small <>. 



(i. Scales smaller, radii more parallel .... Limanda ferruginea Storer. 



Scales larger, radii less parallel 



Hippoglossoides platessoides Fabricius. 

 The last two are nearly alike. 



C. Scale cycloid on both sides (Pleuronectidse). 



Scales very small, long-oval, basal radii few 



Glyptocephalus cynoglossus L. 

 Albatross Sta. 2546. 



Scales much broader, basal radii many 1. 



1. Scales smaller; those of lower side about half size of those on 



upper Lophofjsetta maculata Mitch. 



Scales larger; those of lower side not so much smaller than those 

 of upper Paralichthys oblongus Mitch. 



The Glyplocephalus scale is the most modified (degenerate) of the whole 

 series. Some Achirinre (not seen by me) are scaleless. With regard to 

 the subfamilies of Pleuronectidse, it will be noted that each subfamily has 

 produced a genus with cycloid scales on both sides. 



It is is obvious that the flat-fishes are descended from fishes with the 

 typical Acanthopterygian (ctenoid) type of scales, and that the cycloid 

 scales have arisen through loss of the ctenoid character. From all we 

 knowabout inheritance, it appears practically certain that this character, 

 once completely lost on the lower side, would not return in the well- 

 developed condition in which it occurs in the soles. It must, therefore, 

 be held that the Soleidfe are not derived from the ordinary Pleuronectidse, 

 but rather from some ancient type of flat-fish, probably now without liv- 

 ing representatives. There is, however, a living genus of Pleuronectids 



