220 
COLOUR IN NATURE 
CHAP- 
outgrowths, and that in fishes in general the epidermis 
is little specialised ; in accordance with this we find 
that the epidermis is usually almost entirely devoid 
of pigment, or when, as in the flounder (.Pleuronectes 
flesus ), some pigment is present, it has no effect on 
the visible coloration. In birds, on the other hand, 
where the epidermis gives rise to greatly specialised 
outgrowths, the feathers, the pigment of the body 
is found in these epidermal structures. There is, 
however, a general consensus of opinion that in all 
cases the pigment originates either in the dermis or 
in deeper connective tissue cells and migrates out¬ 
wards to the epidermis, in the cases where it is 
found there. Many would regard this as a proof 
that pigments are in essence waste products, and 
that this migration outwards is to be interpreted as 
a process of excretion, the connective tissue cells or 
amcebocytes carrying the waste pigment from the 
essential organs of the body outward to the inert 
skin. For an elaborate hypothesis of this kind the 
reader may be referred to a paper by Mr. H. E. 
Durham; it is doubtful, however, whether such 
generalisations are not as yet premature. 
To return from this digression to the immediate 
subject of the skins of fishes, we find that, accord¬ 
ing to Cunningham and M'Munn, the elements im¬ 
portant in coloration occur in two layers in the skin, 
and the two layers differ considerably in the upper 
and lower surfaces of the fish. In most cases the 
outer layer occurs in relation to the scales, while the 
inner layer lies close to the surface of the muscles, 
but in the mackerel the loose scales contain no 
colouring elements. The outer layer consists of 
