Doe EXPERIMENTS AND OBSERVATIONS 
The direct resistance which prevents the fish from falling under the 
action of gravity is, of course, friction ; but unless there was some force 
to press the fish against the solid surface there would be no friction. 
In testing the currents with carmine it was seen to pass in at the 
mouth and out at both gill openings and nothing was observed which 
indicated a special part played by the perforation of the gill- 
septum. 
To test the validity of the explanation I had arrived at concerning 
the adhesion of the fish, I constructed a model in the following 
manner. I took a rectangular piece of flexible thin sheet india 
rubber. In the middle of each of the shorter sides I fastened, by a 
few stitches of sewing cotton, a short piece of glass tubing. On to 
one of these pieces I adjusted a long piece of rubber tubing. I placed 
the apparatus in a tank, bringing the tubing out over the edge of the 
olass front, and allowing it to act as a siphon, drawing water out of 
the tank, While the siphon was running I placed the india-rubber flap 
cently against the glass inside the tank under water, and it remained 
adhering to the surface. ‘The front piece of glass tube now repre- 
sented the respiratory channel of the fish, and above and below it 
were apertures between the front of the tank and the rubber flap, re- 
presenting the apertures above and below the snout in the case of the 
fish. The action of the siphon represented the pumping action of the 
muscles and fins in the fish. There was nothing in the model to 
represent the respiratory movements of the fish, but that does not 
invalidate the comparison. When the siphon was stopped by pinching 
the rubber tubing outside the tank, the rubber flap fell away from 
the glass and sank slowly to the bottom of the tank. I think with 
this confirmation the evidence I have now given in support of my 
explanation of the adhesion of the fish to vertical surfaces is amply 
sufficient. I have observed that when other kinds of flat-fish cling 
to the vertical surface of the glass front of an aquarium they move the 
posterior parts of the unpaired fins in the same way as Zeugopterus, 
but these parts of the fins and the muscles adjacent being less deve- 
loped, the action is neither so powerful nor so long maintained. 
There can be little doubt that the explanation above given applies 
to Z. unimaculatus, in which the habit of adhesion was observed by 
Brook, as well as to Z. punctatus. It probably applies also to Z. 
norvegicus, but I have not yet ascertained whether this species has 
the same habit. We do not know at present whether there are 
any other important differences in mode of life between Zeugopterus 
and other genera,—such, for instance, as kind of food. We know 
that these flat fishes are not entirely confined to rocks, but are also 
found on sandy bottoms with other flat fishes. There is no doubt 
that when adhering to a rock, one of these fishes accommodating 
