D.—ZOOLOGY. 85 
all of you do not know that it has been shown that almost as soon as the 
fish is hatched the cartilaginous supraorbital bar above the eye that is 
_ going to migrate begins to be absorbed, and is eventually represented only 
by short processes of the otic and ethmoid cartilages, with a wide gap 
between them; through this gap the eye migrates, with the result that 
when ossification begins the main part of one frontal bone is on the wrong 
side of its eye. The flat-fishes are an offshoot of the perch group, and 
it is known that some of these have a habit of resting on one side ; if such 
a fish found it profitable to lie in wait for its prey in this position, it would 
naturally try to make some use of the eye of the under-side, pressing it 
upwards against the edge of the frontal bone. And in the flat-fishes the 
migration of the eye into and across the territory of the frontal bone, 
prepared for by the absorption of the cartilaginous precursor of the frontal 
bone before the eye shows any sign of migration, may well be interpreted 
as the final stage of a process thus initiated. 
You will have seen, then, that I am inclined to accept Darwin’s theory 
as a whole, including both natural selection and the inherited effects of 
use and disuse, at any rate until some better explanation of the facts is 
forthcoming. But still there are difficulties and to illustrate them I must 
"give one more example from the fishes. 
The most primitive spiny-rayed fishes are the Berycoids, which 
flourished in Cretaceous times; in some of these the vertebrae number 
24, 10 precaudal and 14 caudal. In many families of Percoids, not at 
all closely related to each other, we find this number of vertebre is a 
constant family character ; for example, all the genera and species of Sea- 
breams (Sparide), Red Mullets (Mullide), Cheetodonts (Chetodontide), 
Gray Mullets (Mugilide), and Barracudas (Sphyrenidz) have 24 (10+ 14) 
vertebre. The conclusion is inevitable that this is a primitive Percoid 
‘character derived from a Berycoid ancestor. Yet we have clear evidence 
that whenever the circumstances demanded it this number could be 
decreased or increased. There is no variation and therefore no material 
for selection; also the number of vertebre is settled at a very early stage, 
and no fish can increase or diminish that number in its lifetime. Psettodes, 
the most primitive living flat-fish, has 24 (10+ 14) vertebre ; it is simply 
an asymmetrical perch. It has a large mouth and strong, sharp teeth, 
and its principal movements are probably short dashes after fishes that 
come near enough to be caught. But in other flat-fishes the number of 
vertebree is greater ; in the sole, which feeds on small invertebrates that 
it finds in the sand, and swims along with undulating movements of the 
whole body, the number is about fifty, and in the Tongue-Soles (Cyno- 
glossus) there may be as many as seventy vertebre. ; 
We are almost compelled to believe that muscular movements, the 
efforts of a fish to swim in a certain way, may lead to an alteration in the 
number of muscle segments of its descendants ; the number of vertebre 
is, of course, determined by the number of muscle segments. This is an 
extension of the Lamarckian theory, and some of you may regard it as a 
teleological speculation unworthy of serious consideration ; some may even 
_ think that, as my suggested explanation is incredible, we have here another 
example of the truth of the mutation theory, which in effect states that it 
is only by accident that a structure has a function. 
Many biologists have adopted Weismann’s germ-plasm hypothesis so 
