256 EXPERIMENTS AND OBSERVATIONS 
unpaired fins well separated ; pelvic fins nearly or quite symmetrical ; 
body dextral. Species arctic or subarctic in distribution, 
In this sub-family Jordan and Goss make our common dab the 
type of a genus. ‘The characters are— 
No accessory branch to the lateral line, but a distinct arch in 
it over the pectoral. 
Scales rough, ctenoid, and imbricated. 
Vertebree forty in number. 
The authors recognise four species. One of these is doubtful, 
founded by Prof. Goode on certain specimens taken in deep water 
off the southern coast of New England. Limanda limanda, the 
Kuropean form, has no rugose prominences above the operculum 
behind the interorbital ridge. . The fin-rays are—dorsal 65 to 78, 
ventral 50 to 62; scales along the lateral line 86 to 96, teeth in an 
irregular series. This species extends from the Atlantic coasts of 
France along all the coasts of Northern Kurope, and on the coast of 
Iceland: it is absent from the Mediterranean. 
Iimanda ferruginea is the dab of the American side of the 
Atlantic, extending from New York to Labrador. It differs from 
ours in having more numerous teeth in a more regular close-set 
series, a more projecting snout, and rugose prominences above the 
operculum. ‘The fin-rays are a little more numerous, namely, D. 85, 
A. 62. The scales are smaller and more numerous, namely, lateral 
line 100. 
Timanda aspera is the dab of the North Pacific. It is distin- 
guished by somewhat marked characters, of which the principal are 
that there is no angle between the snout and the profile of the head, 
and the scales of the blind side are more or less rough, those of the 
upper side rougher than in the other species. Specimens have been 
taken on the coasts of both Alaska and Kamtschatka. It seems, 
therefore, that while the species on opposite sides of the Atlantic 
are different, those on opposite sides of the Pacific are the same. 
This case offers an instance of geographical races. The differences 
are not great, but if they are constant it matters lttle whether we call 
these forms, species or varieties or races. ‘The case affords a contrast 
to that of Zeugopterus ; in the latter we have three species in the 
same area, separated by no barriers except those which are physio- 
logical: in other words, they do not interbreed. In the case of the 
three forms of Limanda interbreeding is physically impossible, 
except where the ranges meet. We have no evidence that the 
differences are adaptational. 
Closely allied to Limanda is a species in the Pacific called by 
Jordan and Goss Lepidopsetta bilineata. Only one species is placed 
in the genus, the establishment of which seems superfluous, ‘The 
