Mechanism of reflexion in silvery layers of fish and cephalopods

The highly reflecting structures found in the integuments and eyes of fish and cephalopods were studied. In all cases they consist of alternate layers of high and low refractive index (n) material (in fish, guanine and cytoplasm) and the high refractive index material is in the form of discrete plates. The highest reflectivity at a given wavelength $\lambda _{0}$, together with the widest waveband of high reflectivity, would be given if these alternate layers all had an optical thickness of $\frac{1}{4}\lambda _{0}$. We have examined the possibility that fish and cephalopods can make 'ideal' reflectors of this kind. The thicknesses (t) of the discrete plates released by scratching the reflecting layers were measured by interference microscopy, and it was found that although the plates from a region of a given colour sometimes varied greatly in surface area, their thicknesses were approximately constant. With one exception the optical thicknesses (nt) of the plates found in all the structures studied were between 100 and 200 nm. Many of the reflecting structures are highly coloured and in these there was almost always a good correlation between the wavebands best reflected and four times the optical thicknesses of the plates which they contained. The most ventral scales in the juvenile sprat were studied in some detail. At normal incidence these scales reflect best a waveband around 720 nm and the guanine crystals which they contain all have optical thicknesses close to one quarter of this wavelength. The changes in colour with angle of viewing, and with changes in osmotic concentration of the medium in which these scales are placed, support the idea that the spaces between the crystals are $\frac{1}{4}\lambda _{0}$ spaces. These scales have a high reflectivity to the lights penetrating best into the sea at the oblique angles of incidence from which the strongest intensities of daylight fall in life. Qualitative observations on scales from herring and from other regions on the sprat supported the hypothesis that their guanine crystals were also arranged approximately in $\frac{1}{4}\lambda _{0}$ stacks. Similar conclusions were reached for coloured surfaces found in the skin of the horse mackerel, the iris of the neon tetra, the reflecting tapeta of the dogfish and the spurdog, and in the eye of the squid, Loligo forbesi. In the scales of the roach, Rutilus rutilus, the crystals are of thicknesses which indicate that if in 'ideal' $\frac{1}{4}\lambda _{0}$ stacks they would at normal incidence appear red and at oblique incidence green, whilst in fact they are very little coloured. The crystals from the eyes of Callionymus lyra had 4nt around 300 nm (in the u.v.) yet the reflexions given by piles of these crystals were bronze coloured. Possible explanations of these facts are given. In cephalopods the high refractive index plates are lozenge-shaped, flexible and of refractive index about 1$\cdot $56.