D.—ZOOLOGY. 83 
and more active fishes, essentially similar in structure to the modern 
Tarpon, which chases and devours the schools of small fry of other fishes. 
n relation to these more active habits the lobes of the caudal fin became 
longer and more divergent and the rays of the upper lobe, which in the 
Pholidophoridz were supported by the centra and hzmal spines of the 
upturned end of the vertebral column, acquired a firmer support by the 
enlargement and welding together of the neural spines of some of these 
vertebre, which replaced functionally and caused the disappearance of 
the upturned centra. This structure, thus and for this purpose first 
acquired by the Tarpon-like fishes, persists in all the multitude of modern 
fishes derived from them, whatever their habits, whatever the size and 
shape of their caudal fin. And in comparing the herrings with the Pholi- 
dophoridz we see that the difference in caudal structure is due to the 
arpon ancestry of the herrings. 
At the present day the perch-like fishes are dominant in the coastal 
waters of tropical and subtropical seas. One might have thought that 
when the anterior rays of the dorsal fin had become strong, sharp spines, 
weapons of attack and defence, further modifications would be unlikely ; 
but in different offshoots of the perch tribe many extraordinary modifica- 
tions of the spinous dorsal fin occur. In the flat-fishes, where undulating 
movements of the whole dorsal fin are required, the spines have been 
reconyerted into jointed, flexible rays; in the angler-fishes the spinous 
rays have become flexible and the first has moved on to the snout and 
has been modified into a line and bait. In the sucker-fishes the spimous 
dorsal fin has been transformed into a transversely laminated suctorial 
isc placed on the upper surface of the head; when this disc is applied 
to the skin of a shark or of some other marine animal, the lamine, or 
modified fin-rays, are erected and a series of vacuum chambers is formed 
between them. I put forward this example of the sucker-fishes (Echenevs) 
us one that can be interpreted only on the assumption that a change of 
habits preceded a change of structure. In swift-swimming pelagic fishes 
the spines of the dorsal fin are either short or slender and can be depressed 
within a groove so as not to impede rapid motion through the water ; 
the pilot-fish is a fish of this type that has the habit of associating with 
tks. Some similar fish might have found that a spinous dorsal fin with 
s structure could be used for adhesion if the margins of the groove 
e pressed against the skin of the shark and the spines were slightly 
ected; the habit of trying to adhere once established, the evolution of 
the suctorial disc would follow. 
Throughout, the evolution of fishes illustrates the same principles. 
whi anges of structure have been intimately related to, and may even be 
d to have been determined by, changes of habits, and especially changes 
of food and of feeding habits. Evolution has been adaptive, but modifica- 
ions of structure that were originally adaptive persist when they are 
0 longer; they become historical and the basis for further adaptive 
fications. Iam satisfied that these principles, which I have illustrated 
xamples from the group I have specially studied, have a general 
cation. 
@Q 
