248 EXPERIMENTS AND OBSERVATIONS 
investigators. But among the various methods employed there is 
one which has seldom, if ever, been followed—that of surveying the 
various characters of different grades,—specific, generic, family, &c. 
—in order to find whether it is possible to trace a connection between 
them and the habits of the animals which exhibit them, and gene- 
rally to consider how far the principles which have been suggested 
in explanation of the evolution of species are applicable to the 
diagnostic characters of a particular family. On the present occa- 
sion I propose to consider the family of flat fishes. This Journal is, 
I think, not the place for a paper entirely devoted to developing 
arguments or presenting evidence in support of particular views or 
conclusions of a theoretical character. Therefore, although my 
inclination to certain views may be obvious in the following remarks, 
I have no desire to press these views on this occasion ; but my object 
is merely to describe certain observations and studies I have recently 
made, and to point out what an immense field of interesting inquiry 
is afforded in the relations and development of those characters by 
which the subdivisions of a single family of fishes are distinguished. 
My own observations have been for the most part confined to 
British specimens of the family of flat fishes; and for a general 
survey of the characters throughout the family, and their relations 
to one another, I shall rely chiefly on a valuable paper by the 
American ichthyologists, Jordan and Goss, published in 1889.* 
Certain kinds of flat-fish are distinguished by the fact that the 
dorsal and ventral fins are prolonged on to the lower side of the 
body at the base of the tail, the attachments of these accessory 
portions being transverse to the axis of the body. One of the fish 
that present this character is not uncommon round all the coasts of 
Britain ; at Plymouth specimens are frequently obtained, either in 
the Sound in summer, or on more distant grounds. This is the 
Zeugopterus punctatus of Collett, the Muller’s topknot of Couch, 
Rhombus hirtus of Yarrell. The chief other characteristics of this 
fish are its almost rectangular shape, the posterior region of the 
body being much broader and less triangular than in other flat 
fishes, the roughness of the upper side of the body, due to the 
character of the scales, and the presence of a large foramen in the 
septum between the gill-cavities. The great breadth of the body 
posteriorly is due partly to the breadth of the body proper, partly to 
that of the dorsal and ventral fins, in which the fin-rays are longest 
near the posterior end, so that the outer edges of the fins form a 
straight line transversely across the base of the tail. The snout is 
obtuse, and the trunk and dorsal fin rise steeply behind it, giving 
* A Review of the Flounders and Soles (Pleuronectidx) of America and Europe, by 
David Starr Jordan and David Kop Goss; Rep. U.S. Fish Commission for 1886 (1889). 
