Optical fringe effects at prism borders in human tooth enamel sections

SUMMARY When ground sections of mammalian teeth are viewed under the light microscope each enamel prism appears to be surrounded by a prism sheath about 0.5 μm] wide. Seen by the electron microscope these sheaths are at most 0.1 μm wide. Because this dimension is beyond the resolving power of the light microscope it is difficult to explain the origin of the optical image. Optical fringes are frequently seen at the borders of transversely cut enamel prisms. The widths of these fringes have been measured under different optical conditions. It is concluded that the fringes are produced due to interference between direct light rays and those reflected from prism borders. A requisite for clear fringe production is that the reflectance of the prism border should be high. This high reflectance could only be achieved if prisms are separated from each other by material having a refractive index which differs significantly from that of the prisms themselves. It is therefore suggested that prisms are separated by a protein layer. It is then possible to explain why prism sheaths appear 0.5 μm wide under the light microscope. Study of mammalian tooth enamel by light microscopy suggests to the observer that the 5 μm wide rods (generally referred to as prisms) which constitute the enamel are each partially surrounded by a ‘sheath’ about 0.5 μm wide. Micro‐radiographs indicate that the sheath is considerably less mineralized than the body of the prism (Glas, 1965). In contrast, from direct observation of enamel sections by electron microscopy, it has been concluded that no prism sheath exists and that prisms are separated by an interface which is bordered on its two sides by differently orientated crystals (Helmcke 1960, 1967; Ronnholm, 1962; Meckel, Griebstein & Neal, 1965). Between these two extremes are some electron microscope observations appearing to demonstrate the presence of a 0.1 μm wide membrane‐like prism sheath devoid of inorganic crystals (Frank & Nalbandian, 1967) or an irregular region of microporosity where the different orientation of adjacent crystals results in a packing defect (Johnson, 1967). Any enamel section observed with the electron microscope may well have been distorted during its preparation for examination. Shrinkage due to dehydration is one of the most likely artifacts. Because the amount of the distortion cannot be known with accuracy it is difficult to decide which of the above electron‐microscope appearances is closest to the actual structure of the prism boundary during life. It has recently been observed that when viewed end‐on, the borders of prisms frequently appear as light striations (Osborn, 1968 and in press). In the present investigation the widths of these light striations have been measured under different optical conditions on an enamel section which was maintained in a hydrated condition. It was thought that this data on the optical properties of the hydrated prism boundary could be used to predict the structure of the boundary in vivo . A similar shrinkage is to be expected in sections examined with the light microscope because these have usually been dehydrated prior to mounting. It is therefore reasonable to try to relate the 0.5 μm wide optical image seen with the light microscope to the two different structures observed with the electron microscope. If no prism sheath exists it might be argued that the refractive index variation between the borders of adjacent prisms could in some way account for the production of the optical image. However, because this variation is probably less than 0.02 (Fremlin & Mathieson 1964) it is not clear how it could be used to explain the high contrast of the optical image. Furthermore, even if a 0.1 μm wide prism sheath exists, it is significantly thinner than the wavelength of visible light; again it is not clear how its presence results in the production of an image in the light microscope.